<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by John Snow Labs on Medium]]></title>
        <description><![CDATA[Stories by John Snow Labs on Medium]]></description>
        <link>https://medium.com/@JohnSnowLabs?source=rss-3251f80c528f------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/0*nSVHrl-6KHy3GNq3.jpg</url>
            <title>Stories by John Snow Labs on Medium</title>
            <link>https://medium.com/@JohnSnowLabs?source=rss-3251f80c528f------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Tue, 17 Nov 2026 10:11:24 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@JohnSnowLabs/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[John Snow Labs Announces the NLP Lab: Free No-Code AI for Domain Experts]]></title>
            <link>https://medium.com/@JohnSnowLabs/john-snow-labs-announces-the-nlp-lab-free-no-code-ai-for-domain-experts-aef5d2eadf46?source=rss-3251f80c528f------2</link>
            <guid isPermaLink="false">https://medium.com/p/aef5d2eadf46</guid>
            <category><![CDATA[john-snow-labs]]></category>
            <dc:creator><![CDATA[John Snow Labs]]></dc:creator>
            <pubDate>Tue, 10 Jan 2023 10:52:57 GMT</pubDate>
            <atom:updated>2023-01-24T16:29:01.795Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/300/1*qXNFvd7kPHtTF3ojMqtObw.jpeg" /></figure><p>Over the last year, Annotation Lab has grown to be much more than a document annotation tool. It became a full-fledged AI system, capable of testing pre-trained models and rules, applying them to new datasets, training, and tuning models, and exporting them to be deployed in production. All those features together with the new Playground concept presented in the current release notes contributed to the transformation of the Annotation Lab into the<strong> NLP Lab</strong>. A new <strong>Playground</strong> feature is released as part of the <strong>NLP Lab’s Hub</strong> that allows users to quickly test any model and/or rule on a snippet of text without the need to create a project and import tasks. <strong>NLP Lab</strong> also supports the training of Legal and Finance models and Model evaluation for classification projects. As always, the release includes some stabilization and bug fixes for issues reported by our user community. Below are the details of what has been included in this release.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*a79quFpCGCKqnJlP.gif" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*4X5dmzDkVRtUTcs2.gif" /></figure><p>Deployment of models and rules is supported by floating and air-gapped licenses. Healthcare, Legal, and Finance models require a license with their respective scopes to be deployed in <strong>Playground</strong>. Unlike pre-annotation servers, only one <strong>Playground</strong> instance can be deployed at any given time.</p><h3>Training and Preannotation with Finance and Legal Models</h3><p>With this release, users can perform training of Legal and Finance models depending on the available license(s). When training a new model in the NLP Lab, users have the option to select what library to use. Two options were available up until now: Open source and Healthcare. This release adds two new options: Legal and Finance. This helps differentiate the library used for training the models. The new options are only available when at least one valid license with the corresponding scope is added to the License page.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/835/0*k7AUm13jDjkfWuD0.png" /></figure><h3>Publish Trained Models to S3</h3><p>The <strong>NLP Lab</strong> also allows users to easily export trained models to a given s3 bucket. This feature is available on the <strong>Models</strong> page under the <strong>Hub</strong> tab. Users need to enter the s3 bucket path, s3 access key, and s3 secret key to upload the model to the s3 bucket.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/800/0*33f7u7DqhES_1ll6.gif" /></figure><h3>Getting Started is Easy</h3><p>The NLP Lab is a free tool that can be deployed in a couple of clicks on the AWS and Azure Marketplaces, or installed on-premise with a one-line Kubernetes script. Get started here: <a href="https://nlp.johnsnowlabs.com/docs/en/alab/install">https://nlp.johnsnowlabs.com/docs/en/alab/install</a></p><p><em>Originally published at </em><a href="https://www.johnsnowlabs.com/john-snow-labs-announces-the-nlp-lab-free-no-code-ai-for-domain-experts/"><em>https://www.johnsnowlabs.com</em></a><em> on January 10, 2023.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=aef5d2eadf46" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[John Snow Labs Partners with UiPath to Create an Integrated Solution for Data Extraction using…]]></title>
            <link>https://medium.com/@JohnSnowLabs/john-snow-labs-partners-with-uipath-to-create-an-integrated-solution-for-data-extraction-using-8c4f42318f4b?source=rss-3251f80c528f------2</link>
            <guid isPermaLink="false">https://medium.com/p/8c4f42318f4b</guid>
            <category><![CDATA[natural-language]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[data-extraction]]></category>
            <category><![CDATA[naturallanguageprocessing]]></category>
            <dc:creator><![CDATA[John Snow Labs]]></dc:creator>
            <pubDate>Tue, 22 May 2018 14:24:20 GMT</pubDate>
            <atom:updated>2018-05-23T06:09:37.467Z</atom:updated>
            <content:encoded><![CDATA[<h3>John Snow Labs Partners with UiPath to Create an Integrated Solution for Data Extraction using Natural Language Understanding</h3><blockquote><strong>The partnership will give organizations the capability to extract sophisticated information from large, free-text documents using an efficient, unified user interface</strong></blockquote><p>John Snow Labs is hugely proud to announce a partnership with <a href="https://www.linkedin.com/company/uipath/">UiPath</a>, the leading enterprise Robotic Process Automation (RPA) platform, to deploy an integrated and sophisticated fact-extraction technology. The announcement will be officially made at the <a href="https://conferences.oreilly.com/strata/strata-eu/public/schedule/detail/68625">Strata Data Conference</a> in London, one of the world’s most important big data conferences, where the companies will present their integrated data extraction solution on May 24th.</p><p>John Snow Labs helps accelerate the work of data scientists thanks to its turnkey AI platform and AI implementation services. The company’s advanced natural language processing (NLP) library and deep-learning models will be integrated with UiPath’s RPA technology, providing organizations with a complete software platform to automate business processes that depend on understanding human language.</p><p>The companies’ integrated solution will allow users to extract information from free-text documents, often in excess of a hundred pages, saving countless hours of labor. RPA technology is changing businesses around the world at a staggering clip. <a href="http://deloitte.wsj.com/cio/2016/07/06/robotic-process-automation-slashes-it-costs-alleviates-complexity/">According to Deloitte</a>, a project that might cost a company $10 million can now be done using RPA for as little as $5,000 or $10,000. In many cases, combining RPA and NLP technologies exceeds even the most practiced experts in analyzing a variety of materials including legal documents, financial filings, medical records, news stories and live conversations. Delivering this technology in an easy-to-use platform opens up a broad range of new applications across countless industries.</p><p>Traditional rule-based or expression-matching techniques are effective at extracting information from simple, templated documents. But when the information is less straightforward, these techniques often fall short. Distilling accurate information when statements are implied or absent, or in the combination of other statements requires state-of-the-art deep learning techniques applied to NLP.</p><p>UiPath’s workflows and activities for document import/export, optical character recognition, and taxonomy mapping will combine with <a href="http://nlp.johnsnowlabs.com/">John Snow Labs’ NLP library</a> for extracting facts or fields that require non-trivial language understanding. Users will be able to leverage this capability within a single, unified user interface. They’ll also be able to incorporate customized NLP models specific to their industry and document types.</p><p>“Integrating with the UiPath platform enables us to deliver real value on a cutting-edge natural language understanding project within a few weeks, by complementing our NLP technology with the <a href="/service/mailto:https//www.uipath.com/platform">best RPA platform</a> around,” said <a href="https://www.linkedin.com/in/saif-addin-ellafi-74679927/">Saif Addin Ellafi</a>, lead NLP engineer at John Snow Labs. “This provides our customers with a complete, out-of-the-box solution.”</p><p>“UiPath is excited to support this technology partnership and support a seamless integration of John Snow Labs’ state-of-the-art NLP technology inside UiPath Activities,” said <a href="https://www.linkedin.com/in/geroth/">George Roth</a>, Senior Manager for partnerships and alliances at UiPath. “The joint capability is already providing value to business customers and is broadly applicable.”</p><p>The partnership between John Snow Labs and UiPath merges two thriving global industries. The data science platform market is growing at a remarkable rate, with an expected value of <a href="https://www.datascience.com/blog/what-is-a-data-science-platform">$101.4 billion</a> by 2021. The RPA market is expected to <a href="https://www.statista.com/statistics/740440/worldwide-robotic-process-automation-market-size/">exceed $2 billion</a> the same year. Together, John Snow Labs and UiPath will provide creative solutions to critical business issues such as dirty data, understaffing and a lack of management support.</p><p><strong>About John Snow Labs</strong></p><p>John Snow Labs Inc. is an award-winning data operations company, accelerating progress in data science and analytics by taking on the headache of managing data and platforms. A third of the team have a PhD or MD degree and 75% of team members have at least a Master’s, coming from multiple disciplines covering data research, data engineering, data science, security and compliance. A Delaware Corporation, John Snow Labs runs as a global virtual team located in 16 countries around the globe. The company believe in being great partners, in making customers wildly successful, and in using data philanthropy to make the world a better place. Learn more about John Snow Labs on <a href="http://www.twitter.com/johnsnowlabs">Twitter</a>, <a href="http://www.linkedin.com/company/johnsnowlabs">LinkedIn</a>, or our <a href="http://www.johnsnowlabs.com/dataops-blog/">blog</a>.</p><p><strong>About UiPath</strong></p><p>Built for both business and IT, <a href="/service/mailto:www.uipath.com">UiPath</a> is the leading platform for enterprise Robotic Process Automation (RPA). More than 1,100 enterprise customers and government agencies use UiPath’s Enterprise RPA platform to rapidly deploy software robots that perfectly emulate and execute repetitive processes, boosting business productivity, ensuring compliance and enhancing customer experience across back-office and front-office operations.</p><p>With a thriving RPA developer community of more than 120,000 worldwide, UiPath is on a mission to democratize RPA and support a digital business revolution. Based in New York City, UiPath’s presence extends to 14 countries throughout North America, Europe and Asia. The company is backed by $183M in series A &amp; B funding from Accel, CapitalG, Kleiner, Perkins, Caufield &amp; Byers, Credo Ventures, Earlybird’s Digital East Fund and Seedcamp. UiPath’s Series B valuation exceeded $1 billion.</p><p><em>Originally published at </em><a href="https://www.johnsnowlabs.com/dataops-blog/john-snow-labs-partners-with-uipath-to-create-an-integrated-solution-for-data-extraction-using-natural-language-understanding/"><em>www.johnsnowlabs.com</em></a><em> on May 22, 2018.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=8c4f42318f4b" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[GDPR and Data Platforms]]></title>
            <link>https://medium.com/@JohnSnowLabs/gdpr-and-data-platforms-bb6d86d1f32b?source=rss-3251f80c528f------2</link>
            <guid isPermaLink="false">https://medium.com/p/bb6d86d1f32b</guid>
            <category><![CDATA[gdpr]]></category>
            <category><![CDATA[data]]></category>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[data-curation]]></category>
            <dc:creator><![CDATA[John Snow Labs]]></dc:creator>
            <pubDate>Wed, 09 May 2018 11:04:03 GMT</pubDate>
            <atom:updated>2018-05-10T05:43:24.785Z</atom:updated>
            <content:encoded><![CDATA[<p>The report from the House of Lords Select Committee on Artificial Intelligence is 183 pages long. To quote from it, “All indications suggest that the present status quo will be disrupted to some extent by the upcoming General Data Protection Regulation (GDPR), which the UK is planning to adhere to regardless of the outcome of Brexit, and the new ePrivacy Regulation, which both come into force across the EU from 25 May 2018. With respect to data access, the GDPR’s introduction of a right to data portability is probably the most significant feature.”</p><p>Despite heavy criticism of the firm’s track record, Facebook creator Mark Zuckerberg has promised the United States Congress to implement controls similar to GDPR across the globe.</p><p>Internet creator Sir Tim Berners-Lee proposed Solid (a term derived from “social linked data”), a set of standards and tools that would allow internet users to choose where to keep their personal data and how it should be used.</p><p>At the other end of the spectrum, Sensetime is the technology partner enabling China’s nationwide facial recognition program.</p><p>Regulators are focusing on “data monopoly” to escape the “data poverty” bucket. These changes benefit the end user in addition to data control over ownership.</p><h3>Source of Truth: Federated vs. Central</h3><p>To borrow from Redux, the framework for javascript, three fundamental principles are: (1) single source of truth, (2) state is read-only, and (3) changes are made only via pure functions. The antithesis to this is that silos are typically how Conway’s law manifests in big organizations’ first iteration to agility or the status quo at the beginning of their agility journey. While breaking down silos usually sets in motion a mad rush to migrate to the cloud, this is quickly followed by cloud budgeting. Despite players like Cloudability and Unravel Data, organizations rarely conduct the <em>a priori</em> calculation of post-cloud migration OPEX until well into OPEX redlines post migration. Privacy and governance become step children and side effects in this process.</p><p>Traditional organizations that are not moving to the cloud en masse or who are cherry picking applications or silos to move to the cloud may end up with a federated data infrastructure across on-premise infrastructure, the cloud, and hybrids (public and privately managed clouds). Those organizations that fall in the all-in-the-cloud or on-premise-only do not qualify as centralized although that might be the paradigm they fall into inadvertently.</p><p>As you can see, ending up with federated or centralized data sources is almost never a choice but rather an accident or side effect. Given that, how do you ensure data quality/ governance by design? Where there is no single source of truth, product development best practices that are tools driven are most successful.</p><p>Focusing on single source of truth almost never results in the same C-suite sponsorship as cloud migration, cloud budgeting, or even data quality. Single source of truth is a means to reach some of these funded objectives. Data pipelines are a great way to ensure auditability and reproducibility of data used in any experiment that includes code, data, model, and annotations that go with the experiment.</p><p><a href="http://www.johnsnowlabs.com/">John Snow Labs</a> allows you to white label and use the same platform we’ve been using to provide answers to our clientele’s modeling and algorithm needs with a perpetual license. The advantage is that we are experts acknowledged for our handling of unstructured data alongside other data types. In addition, the platform is designed for environment agnostic-cloud, on-premise, and hybrid systems. Our <a href="https://johnsnowlabs.us10.list-manage.com/track/click?u=dbaf2cf587f6389c36364b613&amp;id=42466a9015&amp;e=9158c409aa">Spark-NLP 1.5 </a>benchmarks have just been published. For NLP questions please join our <a href="/service/mailto:nlp@johnsnowlabs.com">Slack </a>channel.</p><h3>Third Party Data Sets</h3><p>The common refrain on the most encountered bottleneck among data scientists, analysts, and machine learning engineers gravitates around the idea that you are only as good as your data. The best of breed inference algorithms cannot reach their full potential without the training data. Likewise, cross validation cannot reach its accuracy full potential without good data quality. Data enrichment is not dependent on just internal data. You have to augment your data using reliable third party data sources. If your pipeline is dealing in C.diff drug predictions, you may need microbiome and metagenome datasets that augment your data set. John Snow Lab’s team of PhDs and MDs have been generating aggregated datasets at the highest quality for the longest time. Please visit <a href="https://johnsnowlabs.us10.list-manage.com/track/click?u=dbaf2cf587f6389c36364b613&amp;id=462e77a422&amp;e=9158c409aa">our catalog</a> to find data sets that can augment your modeling and algorithm needs. John Snow Labs can also find additional data sets not listed in our catalog for your use.</p><p>The standards relating to Data Quality and Governance that you apply to your first party data for handling, processing, and classification practices should not be different and always must augment those that you apply to third party data (in addition to the third party’s agreement with its first party sources). The primary takeaway from the Facebook and Cambridge Analytica hearings is this that just because you bought or sold the data does not mean that the terms and conditions expire at the time of transaction. This is particularly relevant for confidential health care and life sciences data when aggregated. A case in point is the “Golden State Killer” arrest based on a family member’s DNA. The importance of data handling doesn’t get any more apparent when degrees of separation are involved.</p><h3>Logging, Audit Bots</h3><p>Security, Healthcare, and DevOps practitioners generally choose to log all their data and store it permanently. Other industries prefer to meet a minimum bar by logging as little data as possible. The current state of streaming data and inference on real- time production data has given rise to “Observability” over monitoring. While “testing in production” is still a hot button issue, observability and replay are notions that the data stewards community can actively get behind. Even reproducibility is run serverless these days. No code is okay, but observability is a must.</p><p>Pulling from archived storage to recreate timelines of events is the norm as a business practice rather than a chaos engineering notion. Where IT custodians stop at defining the policy, compliance usually takes over so they meet audit requirements in wall clock time scales.</p><p>Cognitive logging bots that anticipate red flags and keep things in near-term archival is a use case of an effective bot in this space. There is a move back to Solr-based logs in place of elastic search because of time and cost.</p><h3>Black Boxes and Fine Prints</h3><p>System integrators today want their methods to be less opaque and bolder rather than provided in copious binders that are never read. Local Interpretable Model- Agnostic Explanations (<a href="https://www.oreilly.com/learning/introduction-to-local-interpretable-model-agnostic-explanations-lime">LIME</a>) and Open Neural Network Exchange (<a href="https://arxiv.org/abs/1711.00694">ONNX) Open AI</a> are steps in the right direction to unbox the black box. The call for a Hippocratic Oath equivalent amongst data stewards should lead to accountability and that requires all the tooling we can get. There’s more than just a fine as incentive to be acknowledged here with GDPR. This is a chance to get this right with privacy and interpretability as first class citizens. Some popular GDPR tools out there include Collibra and Protegrity that are built on the notions of master data management and single source of truth as cloud native as well as environment agnostic on the last on-premise and hybrid holdouts for many healthcare and life sciences industries.</p><h3>About Us</h3><p><a href="http://www.johnsnowlabs.com/">John Snow Labs</a> Inc. is a DataOps company, accelerating progress in analytics and data science by taking on the headache of managing data and platforms. A third of the team have a PhD or MD degree and 75% have at least a Master’s degree from multiple disciplines covering data research, data engineering, data science, and security and compliance. We are a USA Delaware Corporation, run as a global virtual team located in 15 countries around the globe. We believe in being great partners, in making our customers wildly successful, and in using data philanthropy to make the world a better place.</p><p>Learn more about us on <a href="http://www.twitter.com/johnsnowlabs">Twitter</a>, <a href="http://www.linkedin.com/company/johnsnowlabs">LinkedIn</a>, or our <a href="http://www.johnsnowlabs.com/dataops-blog/">blog</a>. <a href="/service/mailto:john@johnsnowlabs.com">Email</a> us to let us help you on your next cutting edge journey.</p><h3>References</h3><p>Onnx, <a href="https://github.com/onnx/onnx-mxnet">https://github.com/onnx/onnx-mxnet</a></p><p>Collibra, <a href="https://www.collibra.com/">https://www.collibra.com/</a></p><p>Protegrity, <a href="https://www.protegrity.com/">https://www.protegrity.com/</a></p><p>LIME, <a href="https://arxiv.org/pdf/1602.04938v1.pdf">https://arxiv.org/pdf/1602.04938v1.pdf</a></p><p><em>Originally published at </em><a href="https://www.johnsnowlabs.com/dataops-blog/gdpr-and-data-platforms/"><em>www.johnsnowlabs.com</em></a><em> on May 9, 2018.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=bb6d86d1f32b" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Training Data for Machine Learning Algorithms Done Right!]]></title>
            <link>https://medium.com/@JohnSnowLabs/training-data-for-machine-learning-algorithms-done-right-d873c9cc3ed1?source=rss-3251f80c528f------2</link>
            <guid isPermaLink="false">https://medium.com/p/d873c9cc3ed1</guid>
            <category><![CDATA[dataops]]></category>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[data]]></category>
            <dc:creator><![CDATA[John Snow Labs]]></dc:creator>
            <pubDate>Mon, 30 Apr 2018 09:09:28 GMT</pubDate>
            <atom:updated>2018-05-03T05:01:06.224Z</atom:updated>
            <content:encoded><![CDATA[<p><strong>Machine Learning vs. Statistical Modeling</strong></p><p>There are lots of differences between traditional statistical modeling and machine learning. Machine learning depends on computer algorithms, few estimates, and small to large datasets to produce predictions of high accuracy.</p><p>Statistical modeling, on the other hand, depends on human talents, mathematical equations, and many assumptions to predict the “best estimate.”</p><p>Figure 1: Gradient Descent [Source: <a href="https://en.wikipedia.org/wiki/Gradient_descent]">https://en.wikipedia.org/wiki/Gradient_descent]</a></p><h3>How much training data do I need to develop my model?</h3><p>The amount of training data needed to build a comprehensive machine learning model is highly debated. The type of data determines the number of features. The greater the number of features, the greater the amount of data needed (e.g., bioinformatics and genetics models will need more training data than real estate models). The golden rule is that the amount of data should be 10 times the number of features. However, this is also debatable because in some cases, gathering enough data is the real issue (imagine building a model for tsunamis).</p><p>Another factor is the choice of evaluation model. Evaluation models may use the same dataset to train and test or may follow the train/test split methodology.</p><p>John Snow Labs has cleaned and curated over <a href="https://www.johnsnowlabs.com/data-2/">1,800 datasets</a> across 18 areas of human health and well-being. We also have a team of healthcare research specialists at your service. You can check out the John Snow Labs <a href="https://www.johnsnowlabs.com/data-2/">data library here</a>.</p><p>Our top five datasets by number of data points:</p><ol><li>51,350,000</li><li>24,525,000</li><li>18,275,000</li><li>14,375,000</li><li>11,675,000</li></ol><p>KDnuggets™ has a good <a href="https://www.kdnuggets.com/2015/03/more-training-data-or-complex-models.html">blog</a> addressing how much data is needed for more complex models. Practically speaking, the domain experience and plotting a learning curve are the best ways to decide whether you have enough data. These will expose the training progress and test errors while you increase the training dataset size.</p><h3>The Gradient Descent (J) and Linear Regression</h3><p>You need to develop a model to be able to predict the outputs. Assume that you have a function <em>H(X).</em> This function is defined by a set of parameters or coefficients <em>Ɵ0, Ɵ1, Ɵ2, Ɵ3…etc.</em>) and can be represented as:</p><p><strong><em>H0 (X) = Ɵ0 + Ɵ1 X</em></strong></p><p>This equation is a linear equation, meaning that the solution set can be represented by a straight line. The closer the value of <em>H(X)</em> to the value of <em>Y</em> in the coordinates <em>(X, Y)</em> of the training data, the more precise the data model will be. In other words, you need the minimum difference between the estimator (training dataset) and the estimated value (prediction) to determine that you have a good model. This can only be achieved by choosing values for Theta (Ɵ) or for the parameters that can achieve this minimum difference.</p><p>You must use a formula called the “Cost Function [<em>J(Ɵn</em>)]” to calculate the error (use the Mean Squared Error [MSE] function). The MSE function will measure the difference between the training dataset and the predicted values.</p><p>(Cost Function)</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/277/0*NXVAtHt_XzRZYFqK.png" /></figure><p>Your next goal should be to minimize the value of this function (to minimize the error). This can be accomplished by using an algorithm called “Gradient Descent Algorithm.” Gradient Descent is a simple algorithm used for optimization. Use it whenever parameters (coefficients) cannot be calculated analytically (using linear algebra).</p><p>Gradient Descent uses the following function iteratively by assigning an initial value (weight) to the coef<strong>f</strong>icients and then changing these values iteratively proportional to the negative of the function gradient until it reaches convergence. Convergence can be achieved by looking for minor changes in error iteration-to-iteration (where the gradient is close to zero value).</p><p>This equation will be repeated for every (j) until convergence is reached.</p><p><em>Hint: (α) in the equation represents the learning rate. You can initially consider it a value representing how fast you can reach the minimum or simply how much time this process may take.</em></p><p>The efficiency of Gradient Descent with very large training datasets has been questioned. Regardless, machine learning experts can scale Gradient Descent to very large datasets.</p><p>There two main types of Gradient Descent are Batch Gradient Descent and Stochastic Gradient Descent. Batch Gradient Descent can be used with linear and logistic regression, but it is somewhat slow and not suitable for very large datasets. Stochastic Gradient Descent is much faster and more suitable for large training datasets.</p><p>Not only is obtaining training data a challenge, but other challenges include determining how much data you need, cleaning and curating the data, and choosing the best optimization algorithm and evaluation methodology. If you are a data scientist, this will require your valuable time to achieve sound results. Tasks like data curation and choice of methodology can be outsourced to a consultant. The bigger challenge is finding the right experts to assist you. John Snow Labs is the organization you need to keep your (J) at the minima and significantly mitigate the risks of your research project.</p><p><em>Originally published at </em><a href="https://www.johnsnowlabs.com/dataops-blog/training-data-for-machine-learning-algorithms-done-right/"><em>www.johnsnowlabs.com</em></a><em> on April 30, 2018.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=d873c9cc3ed1" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[How Your Company can use Big Data and Analytics to Increase ROI]]></title>
            <link>https://medium.com/@JohnSnowLabs/how-your-company-can-use-big-data-and-analytics-to-increase-roi-8e61ad57b0bc?source=rss-3251f80c528f------2</link>
            <guid isPermaLink="false">https://medium.com/p/8e61ad57b0bc</guid>
            <category><![CDATA[big-data]]></category>
            <category><![CDATA[healthcare]]></category>
            <category><![CDATA[data-science]]></category>
            <dc:creator><![CDATA[John Snow Labs]]></dc:creator>
            <pubDate>Mon, 23 Apr 2018 15:39:28 GMT</pubDate>
            <atom:updated>2018-04-24T05:06:31.255Z</atom:updated>
            <content:encoded><![CDATA[<p><strong>Data Science Informs Successful</strong> <strong>Business Models</strong></p><p>Investments in big data analytics can have a significant impact on your company’s bottom line. The McKinsey Global Institute estimated that companies that inject big data and analytics into their operation show productivity and profitability rates that are 5 to 6 percent higher than those of their peers. Retailers in particular stand to increase their operating margins by more than 60 percent, while the United States healthcare sector could reduce costs by 8 percent through data analytics efficiency and quality improvements (McKinsey Global Institute, 2015). The consensus among healthcare executives is that intelligent machines will provide future industry services on an equal par with qualified healthcare staff and data scientists. Companies that are successful at analyzing data to generate revenue will use data analytics to identify valuable business opportunities and marketing strategies, develop targeted products and services, and effectively deliver those products and services to their customers.</p><p><strong>Nine Ways Big Data Analytics can be applied to Increase Your ROI:</strong></p><p>Consumer demand and market rates can be analyzed using data science to identify prices that will yield the maximum profit in real time and determine when to activate the customer base through promotions. On average, a 1 percent price increase translates into an 8.7 percent increase in operating profits (assuming no loss of volume) (McKinsey Global Institute, 2015). An expert in optimizing prices is Marriott International that has established its Total Hotel Optimization program. Property revenue managers and hotel owners are provided with tools to optimize offerings to frequent customers when they are likely to defect to competitors. The company has also created a revenue opportunity model, which computes actual revenues as a percentage of the optimal rates that could have been charged. That figure has grown from 83 percent to 91 percent as Marriott’s revenue-management analytics has been applied throughout the enterprise (Davenport, 2006).</p><ul><li>Effective Product/Service Delivery</li></ul><p>Healthcare centres have integrated triage algorithms to manage staffing, patient transfers, and patient room assignments. Clinical decisions are accelerated by the availability of each patient’s history of medical tests, laboratory reports, and prescribed medications on one electronic dashboard. Some intensive care units are using analytics to evaluate multiple data streams from patient monitors to predict whether a patient’s condition is likely to worsen.</p><p>A well-known example of a company using data science to reduce costs in service is the United Parcel Service (UPS). In 2008, UPS launched routing software called Orion to calculate the most efficient route for each truck. Using Orion to analyze 250 million address points a day and perform 30,000 route optimizations per minute, UPS confirmed that routes mapped to maximize right versus left-hand turns saves the company $300 to $400 million annually in fuel, wages, and vehicle running costs as well as reduces emissions equivalent to over 20,000 passenger cars (Prisco, 2017).</p><p>Some of the largest company costs are employee salaries. Data science initiatives have the potential to reduce man-hours needed to manually enter and review data. The efficient use of staff resources can be tracked by identifying periods of peak business and staffing demands such as inventory arrivals. In addition, individual employee costs can be compared to their contribution.</p><p>Analytics can provide data on equipment use rate, maintenance, and life cycle to predict future inventory needs and the value of repair versus replacement.</p><ul><li>Conduct Client Loyalty Analyses</li></ul><p>Companies can review customer satisfaction data as well as measures of purchase frequency by target consumer populations. Likewise, medical centres report a higher ROI from initiatives related to their customer relationship management (CRM).</p><ul><li>Predict and Prevent Readmissions</li></ul><p>The United States Hospital Readmission Reduction Program (HRRP) financially penalizes hospitals with relatively high rates of Medicare readmissions for patients with heart attacks, heart failure, and pneumonia. Mount Sinai Hospital in New York recognized that reducing readmissions requires an understanding of the unique drivers of readmission for each patient, such as language, literacy, mental health, and access to social support systems, insurance, housing, and transportation. The Hospital successfully implemented a predictive model <a href="http://apps.himss.org/davies/docs/2012_CaseStudies/mountSinai/InnovationPACT.pdf">called the Preventable Admissions Care T</a>eam (PACT) into its electronic medical record system that triggers interventions and enhanced care coordination for high-risk patients after discharge (HIMSS, 2012).</p><p>Analytics can track fraudulent and incorrect payments.</p><p>Data science allows outlier data to be identified and examined to help develop processes to avoid future mistakes.</p><ul><li>Meet Reporting Requirements Efficiently</li></ul><p>When auditors want to review specific clinical or operational data, the collection of multiple streams of data allows targeted, easy-to-read data reports to be quickly generated with minimal work from staff.</p><p><strong>For the Most Economical Incorporation of Data Science into Your Business Model,</strong> <strong>Use an Insights-as-Service Company</strong></p><p>John Snow Labs provides data analytics services with our own team of skilled data scientists that help you retrieve relevant datasets and insights to grow and support your business. We have the software, tools, and access to external data sources at our fingertips that we manage and update ourselves. By partnering with John Snow Labs you can leverage your existing infrastructure and avoid substantial data science start-up and maintenance efforts.</p><p>For Atigeo, a technology company whose flagship big data analytics product xPatterns™ generates knowledge from all available data to deliver insights, predict outcomes and mitigate risks, data analytics is a necessary priority — but not its first. To reduce the time spent finding, cleaning, formatting, updating, and publishing data for analysis, the team of healthcare data analysts at Atigeo partnered with John Snow Labs to save an average of 4,096 data science man-hours per month! For more details, read the case study <a href="http://www.johnsnowlabs.com/wp-content/uploads/2016/09/DataOps-Case-Study-xPatterns-John-Snow-Labs.pdf">here</a>. To learn how big data analytics can address your unique needs and ultimately increase your business value, <a href="https://ida-johnsnowlabs.youcanbook.me/">contact</a> John Snow Labs today.</p><p><strong>References</strong></p><p>Davenport, T. (2006) Competing on Analytics. <em>Harvard Business Review,</em> [online] Available at: <a href="https://hbr.org/2006/01/competing-on-analytics">https://hbr.org/2006/01/competing-on-analytics</a> [Accessed 16 March 2018]</p><p>Healthcare Information and Management Systems Society (2012) The Mount Sinai Medical Center Selected as 2012 Enterprise HIMSS Davies Award Winner. <em>HIMSS News</em>, [online] Available at: <a href="http://www.himss.org/news/mount-sinai-medical-center-selected-2012-enterprise-himss-davies-award-winner">http://www.himss.org/news/mount-sinai-medical-center-selected-2012-enterprise-himss-davies-award-winner</a> [Accessed 5 March 2018]</p><p>McKinsey &amp; Company (2015) <em>Marketing &amp; Sales: Big Data, Analytics, and the Future of Marketing &amp; Sales.</em> [online] Available at: <a href="https://www.mckinsey.com/~/media/McKinsey/Business%20Functions/Marketing%20and%20Sales/Our%20Insights/EBook%20Big%20data%20analytics%20and%20the%20future%20of%20marketing%20sales/Big-Data-eBook.ashx">https://www.mckinsey.com/~/media/McKinsey/Business%20Functions/Marketing%20and%20Sales/Our%20Insights/EBook%20Big%20data%20analytics%20and%20the%20future%20of%20marketing%20sales/Big-Data-eBook.ashx</a> [Accessed 15 March 2018]</p><p>Prisco, J. (2017) Why UPS Trucks (Almost) Never Turn Left. <em>CNN</em>, [online] Available at: <a href="https://www.cnn.com/2017/02/16/world/ups-trucks-no-left-turns/index.html">https://www.cnn.com/2017/02/16/world/ups-trucks-no-left-turns/index.html</a> [Accessed 15 March 2018]</p><p>Vennaro, N. (2017) Buy (don’t build) healthcare data insights to improve data investment ROI. <em>MedCityNews</em>. [online] Available at: <a href="https://medcitynews.com/2017/03/buy-dont-build-healthcare-data-insights-improve-data-investment-roi/">https://medcitynews.com/2017/03/buy-dont-build-healthcare-data-insights-improve-data-investment-roi/</a> [Accessed 9 March 2018]</p><p><em>Originally published at </em><a href="https://www.johnsnowlabs.com/dataops-blog/how-your-company-can-use-big-data-and-analytics-to-increase-roi/"><em>www.johnsnowlabs.com</em></a><em> on April 23, 2018.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=8e61ad57b0bc" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Rich U.S. Vaccine Data Regularly Updated at the John Snow Labs Library]]></title>
            <link>https://medium.com/@JohnSnowLabs/rich-u-s-vaccine-data-regularly-updated-at-the-john-snow-labs-library-c251a304e432?source=rss-3251f80c528f------2</link>
            <guid isPermaLink="false">https://medium.com/p/c251a304e432</guid>
            <category><![CDATA[data-curation]]></category>
            <category><![CDATA[database]]></category>
            <category><![CDATA[vaccines]]></category>
            <dc:creator><![CDATA[John Snow Labs]]></dc:creator>
            <pubDate>Mon, 16 Apr 2018 16:58:57 GMT</pubDate>
            <atom:updated>2018-04-17T04:57:59.131Z</atom:updated>
            <content:encoded><![CDATA[<p><strong>John Snow Labs</strong> is providing improved access to the <strong>United States Vaccine Database</strong> published by the U.S. National Center for Immunization and Respiratory Diseases (NCIRD) of the Centers for Disease Control and Prevention (CDC).</p><p>Vaccine codes can be confounding even for healthcare professionals. In 1999, CDC published an Implementation Guide for Immunization Data Transactions. A set of two main code types were defined, CVX and MVX. Later, other codes were added such as NDC, CPT, and VIS.</p><p><em>The John Snow LABS fully updated U.S. vaccine database with its 13 included datasets</em> will guarantee instant access to vaccines currently approved by the Center for Biologics Evaluation and Research (CBER) of the U.S. Food and Drug Administration (FDA) and will assist users with the understanding of vaccine codes.</p><h3>Vaccines are indispensable to people’s lives and the sustainability of healthcare systems</h3><p>Millions of child, adult, and elderly lives are saved each year by vaccines. Reducing the burden of disease through immunization contributes not only to an increase in life expectancy but also provides financial benefits at the individual and country level.</p><p><em>The use of vaccines to prevent diseases in children, adults, and the elderly results in fewer medical visits, diagnostic tests, treatments, and hospitalisations, which leads to substantial savings in healthcare costs each year.</em> <em>(Largeron N, et. al., 2015)</em></p><p>However, immunization has been linked to adverse effects that have been the subject of great debate. To ensure traceability of these adverse events, all vaccine products are codified by standardized alphanumeric codes.</p><h3>A variety of vaccine codes are listed in the U.S. Vaccine Database</h3><p>The <strong>John Snow Labs</strong> <strong>U.S. Vaccine Database</strong> is one of the most complete collections of datasets in the data market with regard to the vaccines codes. It includes over 16 codes recognized by NCIRD and FDA.</p><p>The two most important codes are <strong>CVX and MVX</strong>. The CVX code has a numeric format and is used to identify the type of vaccine product used. The MVX code is alphabetic and identifies the manufacturer of a vaccine. Together, these two codes refer to the trade name of a vaccine product.</p><p>In addition to this set of codes, other codes available in the <strong>John Snow Labs</strong> database include:</p><ul><li><strong>CPT </strong>(Current Procedural Terminology) Developed and maintained by the American Medical Association (AMA) through the Healthcare Common Procedure Coding System (HCPCS). Used by insurers to help determine the amount of reimbursement a practitioner will receive for services provided. The CPT code IDs are also available in the database;</li><li><strong>NDC</strong> (National Drug Code) codes. Can be considered the most important type of FDA drug codes and serve as universal identifiers for drug products (including vaccines). NDC related codes found in the database are Use Unit Labeler, Use Unit Package, and Use Unit Product;</li><li><strong>GTIN (</strong>Global Trade Item Number) codes and some linked codes such as Sale GTIN and Use GTIN; and</li><li><strong>VIS</strong> (Vaccine Information Statement) related codes to help identify vaccine products for public users.</li></ul><h3>Mapping vaccines codes</h3><p>In addition to access to different vaccine product codes, code datasets can be mapped to identify vaccine codes from other codes. Most mapping starts using CVX as the primary code for vaccine products.</p><h3>Current vaccine status information</h3><p><strong>The John Snow Labs</strong> <strong>U.S. Vaccine Database</strong> is not limited to codes. It allows users to access the status of each vaccine product. For example:</p><ul><li>Active, inactive, never active, pending, or non-U.S.</li><li>Consideration for classification as a vaccine</li><li>Current manufacturer</li><li>VIS edition</li></ul><h3>Additional descriptions and information available</h3><p>In addition to U.S. vaccines codes and status, other information provided in the John Snow LABS database includes:</p><ul><li>vaccine full name</li><li>vaccine description</li><li>CPT description</li><li>labeler related information</li><li>packaging form</li><li>vaccine group names</li><li>VIS document type</li><li>VIS webpage URL</li><li>last record update</li></ul><p>Each one of the 13 datasets included in the database has its own description. All headers are accompanied by additional information on the dataset.</p><p>This information transforms the original datasets from aggregations of codes and numbers to real reference tools for any interested user.</p><h3>Update guaranteed by John Snow Labs</h3><p>For its <strong>U.S. Vaccine Database</strong>, <strong>John Snow Labs</strong> guarantees its clients and partners a monthly update of all included datasets. This allows access to all new vaccines, their related codes, and current status as defined by relevant authorities including FDA and NCIRD.</p><h3>REFERENCES</h3><p>Largeron, N., Lévy, P., Wasem, J., &amp; Bresse, X. (2015). Role of vaccination in the sustainability of healthcare systems. <em>Journal of Market Access &amp; Health Policy</em>, [online] Volume 3(1). Available at: <a href="http://doi.org/10.3402/jmahp.v3.27043">http://doi.org/10.3402/jmahp.v3.27043</a> [Accessed 12 March 2018].</p><p><em>Originally published at </em><a href="https://www.johnsnowlabs.com/dataops-blog/rich-unstates-vaccine-database-regularly-updated-at-the-john-snow-labs-library/"><em>www.johnsnowlabs.com</em></a><em> on April 16, 2018.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=c251a304e432" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Democratizing AI — Build versus Buy for AIML is not Hobson’s Choice]]></title>
            <link>https://medium.com/@JohnSnowLabs/democratizing-ai-build-versus-buy-for-aiml-is-not-hobsons-choice-6833ba5af8e4?source=rss-3251f80c528f------2</link>
            <guid isPermaLink="false">https://medium.com/p/6833ba5af8e4</guid>
            <category><![CDATA[dataops]]></category>
            <category><![CDATA[machine-learning]]></category>
            <dc:creator><![CDATA[John Snow Labs]]></dc:creator>
            <pubDate>Tue, 10 Apr 2018 09:15:47 GMT</pubDate>
            <atom:updated>2018-04-11T04:38:57.088Z</atom:updated>
            <content:encoded><![CDATA[<p>The Internet was revolutionized by a book store named Amazon and search engines du jour. Google ushered in the storage paradigm Amazon Web Services unleashing the 16 digits-to freedom-to-compute era of 1,000 startups a day.</p><p>Computing is now being pushed to the edge with decisions generated using artificial intelligence (AI). Google Search and Google Ads began using Hadoop as implemented by Yahoo to be consumable on a massive scale. Apache Spark has made the movement of data a tad more palatable from a latency standpoint. Streaming players like Confluent’s Kafka, Flink, and Druid have rendered fog computing a reality. Telco is now more than just a tunnel owner and can create different offers to Northbound and Southbound traffic. Data swamps and big data lakes no longer worry about arrival latencies, but give degrees of caps as service level agreements. They have given way to A/B testing anything imaginable, moving us from a data-centric to model-centric world.</p><p>The same AI (reinforcement learning) that defeated Lee Sedol is Automated Machine Learning (AutoML) that learned and created an intermediate language on its own from the translation of human language A to human language B. It has been democratized and is offered from a class of libraries that are ASIC aware (TPUs, GTX 2080s, D-Wave) from a set of world class companies (Tensorflow, Kaldi, PyTorch, MXNet, CNTK, cuDNN, spaCy, and CoreML from Google, Facebook, Amazon, Microsoft, Nvidia, Salesforce, Apple, etc.).</p><p>With this plethora of democratized AI available, remaining a data-centric enterprise without models is literally leaving money on the table. It used to be that throwing any data away was unthinkable because one might find the gold bullet from mining that data in future decades. But now with online learning, transfer and seq2seq models can help find meaning from data almost momentarily.</p><p>No one firm has a monopoly on the variety of available algorithms, libraries, and frameworks, and it’s almost foolhardy to think one can harness the beast all in-house. Agency specialists with domain knowledge make the difference between the winners and the losers.</p><p><strong>Myth 1: <em>Build it all in-house</em></strong></p><p>Your pedigree can be your strength or weakness depending on how you deploy it. Very often, the technologist is attracted to the shiny new toy on the block while the bottom-line check writer is left scratching their head when they realize the lost dollar value in what is being built. The length of capital expenditure and operating expense debates are often indicators of analysis paralysis. The savvy engineering stakeholder sees the value in dividing large undertakings into tractable milestones tied to business use. This also forces organizations to review their in-house skills matrix and map to the requirements as well as augment efforts with first-class vendors and partners across the milestones. When staff augmentation is based on partnerships, build versus buy ceases to be a debate.</p><p><strong>Myth 2: <em>All the data must be harnessed before building the first model</em></strong></p><p>The classic chicken and egg problem manifests itself when stories are blocked across sprints. The time tested solution is to buildmodular; build in small steps. Data wrangling for modeling is usually about whittling down to that small, meaningful, needle dataset from the big data haystack. Using agile groups that deliver in parallel is one tactic to solve the syndrome of forever data blocked.</p><p><strong>Myth 3:</strong> <strong><em>Learning in production is risky</em></strong></p><p>Building for observability is just beginning to gain a foothold in engineering operations. Web-scale application developers quickly learned that observability had to be a feature for uses such as Parse or Tango. The same applies when taking models to production. In reality, the model is exercised only in production</p><p>Online learning in production is the norm in the current generation of hyper-growth companies like Airbnb, Lyft and Uber; taking web-scale Amazon and Google A/B testing to a new level in the machine learning era. This implies building observability, interpretability, and reproducibility into models as first-class features.</p><p><strong>Myth 4:</strong> <strong><em>Interpretability is undefined</em></strong></p><p>Projects like LIME are initial baby steps into the black box of why a neural network works the way it does.</p><p>As machine learning applications increasingly fall into the realm of everyday regulation, interpretability is a first-class feature for model building. TensorfFlow debugging is another step in the right direction.</p><p><strong>Myth 5: This is a money pit</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/219/0*do-e1rUcUMSWDqYo.png" /></figure><p>Reconciling the AI pipe dream with the realities of a quantifiable return is best achieved by good old agile and lean chunking of projects into measurable units. A goal that goes hand in hand with chunking is systematically killing projects that do not show promise. Common counterexamples are Boston Dynamics robots and autonomous driving. Moonshots have had enough of their arms and legs being chopped off to get them back on track. One gets there by being methodical about measuring progress and resetting goals. This is also why vendors and partners on these journeys are key. Not only do they function to provide a diversity of ideas and solutions, but they also share in the economic burden of the outcomes. When you tie a project to business outcomes and pick partners based on well-defined criteria and process, the money pit issue is definitely put to question if not to bed.</p><p><strong>ABOUT US</strong></p><p><a href="http://www.johnsnowlabs.com/">John Snow Labs</a> Inc. is a DataOps company accelerating progress in analytics and data science by taking on the headache of managing data and platforms. A third of the team have a Ph.D. or M.D. degree and 75 percent have at least a master’s degree in multiple disciplines including data research, data engineering, data science, and security and compliance. We are a corporation based in Delaware, run as a global virtual team located in 15 countries around the globe. We believe in being great partners, making our customers wildly successful, and in using data philanthropy to make the world a better place.</p><p>Learn more about us on <a href="http://www.twitter.com/johnsnowlabs">Twitter</a>, <a href="http://www.linkedin.com/company/johnsnowlabs">LinkedIn</a>, or our <a href="http://www.johnsnowlabs.com/dataops-blog/">blog</a>. <a href="https://ida-johnsnowlabs.youcanbook.me/">Contact us</a> for help on your next cutting-edge journey.</p><p><strong>References</strong></p><p>Tensorflow™ (2018). TensorFlow Debugger, [online] Available at: <a href="https://www.tensorflow.org/programmers_guide/debugger">https://www.tensorflow.org/programmers_guide/debugger</a> [Accessed 9 Apr. 2018]</p><p>Tulio Ribeiro, M., Singh, S., Guestrin C. (2016). Introduction to Local Interpretable Model-Agnostic Explanations (LIME), <em>O’Reilly Media, Inc.,</em> [online] Available at: <a href="https://www.oreilly.com/learning/introduction-to-local-interpretable-model-agnostic-explanations-lime">https://www.oreilly.com/learning/introduction-to-local-interpretable-model-agnostic-explanations-lime</a> [Accessed 9 Apr. 2018]</p><p><em>Originally published at </em><a href="https://www.johnsnowlabs.com/dataops-blog/democratizing-ai-build-versus-buy-for-aiml-is-not-hobsons-choice/"><em>www.johnsnowlabs.com</em></a><em> on April 10, 2018.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=6833ba5af8e4" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The effects of quality improvement level on the ACOs saving rates improvements]]></title>
            <link>https://medium.com/@JohnSnowLabs/the-effects-of-quality-improvement-level-on-the-acos-saving-rates-improvements-58a0100a0a9a?source=rss-3251f80c528f------2</link>
            <guid isPermaLink="false">https://medium.com/p/58a0100a0a9a</guid>
            <category><![CDATA[data-curation]]></category>
            <category><![CDATA[healthcare]]></category>
            <category><![CDATA[data]]></category>
            <dc:creator><![CDATA[John Snow Labs]]></dc:creator>
            <pubDate>Tue, 03 Apr 2018 13:47:03 GMT</pubDate>
            <atom:updated>2018-04-04T05:38:09.610Z</atom:updated>
            <content:encoded><![CDATA[<ul><li><em>A quality score difference of at least 5% at two years interval is more likely associated with an improvement of the saving rate</em></li><li><em>The fact that ACOs can be motivated to deliver healthcare services of low health-related utility to keep and increase the beneficiaries’ satisfaction level can lead to decreasing trend of saving improvement rate over a quality improvement of 5%</em></li><li><em>An incremental increase in quality (probably of 5%) in conjunction with a better control on the provision of low-value health services can be more beneficial to ensure a better correlation of what is delivered to ACO beneficiary population and the needs of the population, and so on saving more.</em></li></ul><p>Following the Patient Protection and Affordable Care Act, known as Affordable Care Act (ACA) and nicknamed Obamacare, which was enacted in 2010 and the Health Care and Education Reconciliation Act, enacted in the same year, the first Accountable Care Organization (ACO) model, the Pioneer model, started in 2012. 32 ACOs enrolled in the first performance year and after a period of follow-up of 5 performance years 9 only 9 ACOs remained in the program in 2016. The Pioneer ACO Model was designed for healthcare organizations that were already experienced in coordinating care across different settings. At the same time, in 2011 the Medicare Shared Savings Program was established with the purpose of developing Accountable Care Organizations (ACOs) and its first cycle started 2012 with its first performance year with 27 of healthcare organizations enrolled, 5 of them participating in the Advance Payment Model, developed with the purpose of to encourage rural and physician-based ACOs to participate in the Shared Savings Program. In January 2017 The Centers for Medicare &amp; Medicaid Services (CMS) announced its seventh cycle with 100 new organizations of a total of 178 for this cycle and the total number of 480 ACOs for 2017. The number of 480 organizations includes the 178 organizations enrolled for the seventh cycle. ACOs are groups of healthcare providers (physicians, hospitals) bound together by an executive authority and that need to include primary care providers because beneficiaries are attributed to ACOs based on their use of primary care services.</p><p>According to The Centers for Medicare &amp; Medicaid Services “the intent of the Shared Savings Program is to promote <em>accountability</em> for a population of Medicare beneficiaries, improve the <em>coordination</em> of FFS items and services, encourage investment in <em>infrastructure</em> and redesigned <em>care processes</em> for high quality and efficient service delivery, and incent higher value care” for ACOs. The three-part aim: better health for individuals, better health for population and lower growth in expenditures are monitored through financial results, quality results and measures of health risks among the beneficiary population. Based on the assessment results at the end of a performance year an ACO can benefit, depending on the beneficiary population size, financial savings and the overall quality score, but regardless of the population health risks level of 10–50% share of savings for ACOs that chose to enroll in Track 1 and of 15–60% for those that enroll in Track 2 of the Shared Savings Program. At the same time, the ACOs that opt for Track 2 and are generating losses over an annual (performance year) limit of 5%, 7,5% and respectively 10% must repay 40–60% share of loss.</p><p>Because the Medicare Shared Savings Program main incentive is the proportion of savings an ACO can gain, which calculation is based on the saving rate and quality score obtained at the end of a performance year, the focus of its leading authority must be on achieving a good saving rate and at the same time a good quality score. The quality score is calculated based on the individual score achieved for measures grouped in 3 quality domains: patient/caregiver experience, care coordination/patient safety and preventive health. CMS added to the shared saving rate the second incentive, based on the Improvement Change Score, calculated as the difference of the score for a measure in a performance year and the score the previous performance year. Based on the level of the Improvement Change Score an ACO can receive up to 4 points for each domain within the maximum limit of points for every domain. The second incentive, being directly linked to the main (financial) incentive implies a greater focus on quality improvement, involving greater patient satisfaction level, improved hospital admission management, improved prevention services management. Among the positive effects should be the limitation of induced consumption of healthcare services, prevention of duplicate services and improvement of disease prevention at an individual level. Because the patients’ expectations and needs are not always the same, the healthcare services providers, in order to keep a good satisfaction level among their healthcare consumers, can be forced to deliver health services with a low utility, in these cases. Giving all these effects the improvements in quality score should be positively correlated with the saving rates improvements between performance years and can be negatively correlated if a great “effort” is made to satisfy the health consumers expectations.</p><p><em>The article aim is to display how the quality improvements or changes from a performance year to another and evolution of saving rates between the same years and if there is a point in quality improvement from which the correlation between the two indicators becomes negative.</em></p><p>In order to compare quality improvements/changes and the saving rates improvements/changes between two sets of data belonging to two performance years, the indicators from 2016 and 2014, of 147 ACOs enrolled during the period 2014–2016, were compared. The reason why 2016 and 2014 data were chosen are: the last data available are for the year 2016, the number of ACOs enrolled and for which the data are available for 2016 and for a previous year, is greater when the comparison year is more close to 2016, 2014 is the last year for which the number of ACOs enrolled in both years (2016 and 2014) with available data for the analyzed indicators is a significant one. The last reason is the fact that by choosing 2014, and not 2015, the differences between years for the two indicators should be more visible.</p><p>The differences between the 2016 and 2014 values, for overall quality scores and saving rates, were sorted by the overall quality scores, the overall quality scores differences being in this way ordered from the smallest to the largest. To assess the relation of the saving rates improvement/change, in 2016 compared to 2014, with the overall quality scores improvement/change in the same period in a way that takes into consideration the degree of quality scores changes, the values of the 147 ACOs were divided into 3 groups. Each group consists of 49 ACOs and can be characterized as follows:</p><ul><li>the group containing the 1stthird of the overall quality scores differences, with the lowest ones</li><li>the group of the 2ndthird of the overall quality scores differences, with the middle values</li><li>the group of the 3rdthird of the overall quality scores differences, with the higher values</li></ul><p>The ACOs group with 1stthird of the quality scores differences values (the lowest values) trend and their corresponding saving rates differences trend, that can be saw in the first two charts, indicate the fact that the trend of the lower quality changes in the performance year (PY) 2016, compared with the PY 2014, are associated with a relatively similar slow increasing trend. In the case of the second group, charts 3 and 4, the middle-quality scores differences values (all positive), with a trend line having the interceptor at 4% (0.04 in the equation), are associated with saving rates differences having a similar increasing trend (line) but with the interceptor at -4% (- 0.04). The last two charts indicate that the highest quality scores differences values are associated with a slow decrease in saving rates differences trend. At the same time, if we consider the 19thvalue (of 34%) an outlier, the decreasing trend will keep existing if we replace this value with one of the others high positive values. What can be observed analyzing only the charts with the saving rates differences is that the increasing trend in the charts 2 and 4 stopped in the last chart (6) and at least after a visual analysis we can say that after quality score difference limit the positive effect of quality improvement on the saving rate will stop. According to the chart 5 trend line interceptor (0.05 in the line equation), this value can be around 5% difference in quality score improvement on two performance years interval. At the same time, because the trend line in the last chart is the only one that is doesn’t intersect the abscissa and placed above the abscissa a can be concluded that above the same limit of 5% difference in quality score improvement is a “safe zone” or a “low risk zone” for saving rates improvements.</p><p>As a preliminary conclusion, a difference in the quality score of more than 5% can ensure positive saving rate difference and that the maximum impact of quality improvement on the saving rate improvement is achieved more likely when the difference in quality score is closer to the value of 5%. This means that over the limit of 5% in quality score improvement the increase is based on the same on a mechanism which has a negative impact on the saving rate (improvement). The already mentioned fact that the healthcare providers can be motivated to deliver services that have a low health utility because of the differences between patients’ expectations and needs, can be one of the mechanisms that are causing healthcare services overconsumption.</p><p>According the study “Changes in Low-Value Services in Year 1 of the Medicare Pioneer Accountable Care Organization Program”1, which compared the “use of low-value services between Medicare fee-for-service beneficiaries attributed to health care provider groups that entered the Pioneer program (ACO group) and beneficiaries attributed to other healthcare providers (control group)”, the reduction in low-value health care services (“services that provide minimal clinical benefit”1) during the first year of the Pioneer Program, was a modest one. The study emphasizes the fact that the Shared Savings Programs doesn’t offer incentives to limit the consumption of this type of healthcare services, which combined with the incentive to promote access to satisfactory health services will not favor the reduction of healthcare services of low health-related utility.</p><p>As a final conclusion, an incremental increase in quality (probably of 5%) in conjunction with a better control on the provision of low-value health services can be more beneficial to ensure a better correlation of what is delivered to ACO beneficiary population and the needs of the population, and so on saving more.</p><p><em>Originally published at </em><a href="https://www.johnsnowlabs.com/dataops-blog/the-effects-of-quality-improvement-level-on-the-acos-saving-rates-improvements/"><em>www.johnsnowlabs.com</em></a><em> on April 3, 2018.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=58a0100a0a9a" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Big Data and Pharma Closed Loop Marketing (CLM)]]></title>
            <link>https://medium.com/@JohnSnowLabs/big-data-and-pharma-closed-loop-marketing-clm-532550fd5296?source=rss-3251f80c528f------2</link>
            <guid isPermaLink="false">https://medium.com/p/532550fd5296</guid>
            <category><![CDATA[closed-loop-marketing]]></category>
            <category><![CDATA[big-data]]></category>
            <category><![CDATA[pharma]]></category>
            <category><![CDATA[sales]]></category>
            <dc:creator><![CDATA[John Snow Labs]]></dc:creator>
            <pubDate>Mon, 26 Mar 2018 15:32:36 GMT</pubDate>
            <atom:updated>2018-03-27T05:37:21.270Z</atom:updated>
            <content:encoded><![CDATA[<p><strong>What is CLM?</strong></p><p>CLM relies mainly on data generated through closed-loop reporting. Closed loop refers to the internal reporting process that takes place between sales teams and marketing specialists after sales teams conduct client visits. Such reports include hints and insights gathered by the sales team to assist marketing specialists in the categorization of their clients and their impact on sales. Analyzing this data may help develop new methods or reveal updates for client data. In addition, reporting may indicate when clients are about to opt-out of the current contract or business deal.</p><p><strong>The Types of Data Required for CLM Systems</strong></p><p>CLM systems label service providers (physicians, dentists, therapists, etc.) as “Accounts.” Accounts may be either personal (e.g., a physician) or institutional (e.g., a physician in an institute). The system retains lists for these accounts. Needless to say, the sales team should update the lists to maintain accuracy. For example, no physician should be listed without their current contact information.</p><p>Every physician represents a client. The impact of this client or their importance to the sales team should be assessed, categorized, and listed in the database. Changes in a provider’s status could include death, travel, or retirement. Some physicians could also be irrelevant from the sales team’s point of view. Pharmaceutical companies should provide data about their sales team and what providers they are assigned to. Pharmaceutical sales jobs are also known for their high turnover rates that require continuous updates to the Salesforce data.</p><p><strong>Challenges Facing CLM Systems</strong></p><p>It is expected that pharmaceutical companies will provide their data in different formats (Excel spreadsheets, Microsoft Access databases, JSON files, etc.). Some data might be missing or duplicated. This may present a real threat to the software house handling the CLM system as the system can import standardized data only. All of these potentially thousands of records need to be standardized and cleaned.</p><p>Data science <a href="https://www.johnsnowlabs.com/data-2/">researchers</a> estimate that cleaning such data may take up to 65% of a data scientist’s valuable time.</p><p><a href="https://www.johnsnowlabs.com/">John Snow Labs</a> has dedicated the last two years to developing United States healthcare and pharmacological industry datasets and recently added datasets for these markets from the United Kingdom.</p><p>Within our <a href="http://www.johnsnowlabs.com/data/">healthcare catalog</a>, you can find over 1,800 datasets across 18 areas of human health and well-being. Whatever your business needs, you can <a href="https://ida-johnsnowlabs.youcanbook.me/">schedule a c</a>onsultation with the marketing team to obtain clean and accurate data fast that exceeds your expectations.</p><p><em>Originally published at </em><a href="https://www.johnsnowlabs.com/dataops-blog/big-data-and-pharma-closed-loop-marketing-clm/"><em>www.johnsnowlabs.com</em></a><em> on March 26, 2018.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=532550fd5296" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[John Snow Labs Catalog Reinforces its US and UK Datasets]]></title>
            <link>https://medium.com/@JohnSnowLabs/john-snow-labs-catalog-reinforces-its-us-and-uk-datasets-e435fe9cf5c7?source=rss-3251f80c528f------2</link>
            <guid isPermaLink="false">https://medium.com/p/e435fe9cf5c7</guid>
            <category><![CDATA[dataset]]></category>
            <category><![CDATA[health]]></category>
            <dc:creator><![CDATA[John Snow Labs]]></dc:creator>
            <pubDate>Tue, 20 Mar 2018 10:43:07 GMT</pubDate>
            <atom:updated>2018-03-22T09:10:03.979Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/543/0*jMOERM9fdjJr7YIu.jpg" /></figure><p>In this release, 32 additional datasets are freshly curated and normalized to further extend and strengthen your research and analytics capabilities as John Snow Lab’s efforts zeroes in on quality and efficiency of data as we expand our <a href="http://www.johnsnowlabs.com/data">dataset </a>catalog.</p><p>You, our clientele from the European Region and The United States can benefit from the additional datasets we have primarily curated on <strong>Population Health. </strong>Since population health statistics help establish health outcomes, determinants of health, population health among others, these datasets will help bridge the gap for decision makers to make insightful programs on population health management that will greatly benefit the healthcare industry and the individual recipients as well.</p><p>Currently, population health is one of the drivers of the prevailing health trends in terms of government and private organization’s policies. So in this 1stMarch Release, 14 datasets were added to our John Snow Labs Population Health catalog; 10 are from EU and 4 are from the US.</p><p>5 of these datasets are from the UK Life Expectancy Data on global burden of disease by cause and by risk factors, UK deaths and age-standardized rates, years of life lost, years lived with disability, all of them presented by causes. 2 datasets talk about NHS hospital admission episodes that are related to drug misuse and another on UK women’s smoking status at the time of their delivery. Then 3 other datasets are about UK mortality data from individuals who are under 75 years old whose deaths are from various diseases, OECD demographic indicators and one dataset that covers world statistics on development assistance for health database that covers data from 1990 to 2016 complete this catalog.</p><p>4 more datasets from Population Health include 1 from the US Bureau of Labor statistics on occupational health and safety that shares data on US injuries and illnesses by industry, 1 on nonmarital childbearing data, and 2 more on US Life expectancy data, which talks about the global burden of disease by cause and risk factors.</p><p>In this release, there are also 7 datasets added to <strong>Medical Cost, </strong>5 of which show data on Physician Fee Schedule Relative Value Unit with all 2018 statistics namely on locality country crosswalk, Medicare Economic Index MEI anesthesia conversion factors, Geographic Practice Cost Index GPCI, Outpatient Prospective Payment System OPPSCAP, and Physician Fee Schedule Relative Value Units. 2 more datasets come from Hospital Treatment Events Costs on Health Resource Group HRG by provider, by specialty, HRG Code, and Location.</p><p>Datasets that showcase data on <strong>Census</strong> has 4 new datasets as well; 1 is on UK mid-2016 population estimates by age, 1 on crime data in Baltimore, the US from 2012 to 2018, and 2 more on US Socioeconomic Indicators with data on unemployment rates by metropolitan area and by the unemployment rate changes over the years.</p><p>Statistics on <strong>Hospitals</strong> were also added; 3 datasets on Hospital Compare General Information were expanded, which discusses the nurse prescribers and private controlled drug prescribers in England, plus the abeyance and dispersal general practitioner codes. <strong>Clinical Measures</strong> has 2 new datasets, 1 on Quality Performance Measures, which has information on the London Borough of Barnet performance reporting for the Fiscal Year 2017–2018 and another one on water quality consumer complaints in The United States.</p><p>Lastly, 2 new datasets were affixed to our John Snow Labs catalog on <strong>Drug Pricing</strong> and <strong>Food</strong> data. Drug pricing has 1 data on prescription Items for the treatment of smoking dependence, and for Food, 1 dataset is available about restaurant inspection score, which has data on food hygiene rating data in Oxford City Council.</p><p><em>Originally published at </em><a href="https://www.johnsnowlabs.com/dataops-blog/john-snow-labs-catalog-reinforces-its-us-and-uk-datasets/"><em>www.johnsnowlabs.com</em></a><em> on March 20, 2018.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e435fe9cf5c7" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>