<?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 Lalsingh Ramavath on Medium]]></title>
        <description><![CDATA[Stories by Lalsingh Ramavath on Medium]]></description>
        <link>https://medium.com/@whoisinside?source=rss-58e13d89250f------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/0*AfX5LJ0V1RodW4CO.jpg</url>
            <title>Stories by Lalsingh Ramavath on Medium</title>
            <link>https://medium.com/@whoisinside?source=rss-58e13d89250f------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Wed, 18 Nov 2026 03:22:44 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@whoisinside/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[Designing Agentic AI Systems for Enterprise Search: From OpenSearch to Intelligent Orchestration]]></title>
            <link>https://medium.com/@whoisinside/designing-agentic-ai-systems-for-enterprise-search-from-opensearch-to-intelligent-orchestration-f17dab9d1376?source=rss-58e13d89250f------2</link>
            <guid isPermaLink="false">https://medium.com/p/f17dab9d1376</guid>
            <dc:creator><![CDATA[Lalsingh Ramavath]]></dc:creator>
            <pubDate>Fri, 03 Apr 2026 17:31:01 GMT</pubDate>
            <atom:updated>2026-04-03T17:31:01.380Z</atom:updated>
            <content:encoded><![CDATA[<h3>Introduction</h3><p>Enterprise search systems have traditionally relied on indexing engines such as OpenSearch or Elasticsearch to retrieve data efficiently. While these systems perform well for structured and keyword-based queries, they often struggle with complex, multi-source, and intent-driven queries.</p><p>In modern enterprise environments, data is distributed across:</p><ul><li>Indexed search platforms</li><li>Live APIs (systems of record)</li><li>External services</li></ul><p>This shift requires moving beyond static query execution toward <strong>intelligent orchestration</strong>.</p><p>This is where <strong>Agentic AI systems</strong> come into play.</p><h3>The Problem with Traditional Search</h3><p>Traditional enterprise search systems face several limitations:</p><ul><li>Data duplication across systems</li><li>Complex synchronization pipelines</li><li>Inconsistent data between indexed and live systems</li><li>Rigid, rule-based query execution</li></ul><p>For example:</p><ul><li>Core business data may reside in <strong>live systems of record</strong></li><li>Historical or analytical data may be stored in <strong>search indexes</strong></li></ul><p>Most systems rely on logic such as:</p><blockquote><em>“If field X is present, run query Y”</em></blockquote><p>This approach does not scale well for modern, dynamic use cases.</p><h3>Introducing Agentic AI</h3><p>Agentic AI enables systems to move from static execution to intelligent decision-making.</p><p>Instead of predefined rules, an agent:</p><ul><li>Understands user intent</li><li>Selects appropriate data sources</li><li>Orchestrates multiple calls</li><li>Combines and refines results</li></ul><p>Typical agent workflow:</p><ol><li>Interpret intent</li><li>Plan execution</li><li>Invoke tools (APIs, search systems)</li><li>Aggregate results</li></ol><p>This transforms search into a <strong>decision-driven system</strong>.</p><h3>Architecture: Intelligent Orchestration Layer</h3><p>A modern enterprise search architecture includes:</p><h4>1. Unified API Layer</h4><ul><li>Accepts user queries</li><li>Normalizes request structure</li></ul><h4>2. Agentic Orchestrator</h4><ul><li>Determines query intent</li><li>Selects execution mode:</li><li>Live API</li><li>Indexed search</li><li>Hybrid</li></ul><h4>3. Data Sources</h4><ul><li><strong>Live Systems (System of Record)</strong> → real-time, authoritative data</li><li><strong>Search Index (OpenSearch)</strong> → fast retrieval and filtering</li></ul><h4>4. Response Normalization</h4><ul><li>Merges results</li><li>Ensures consistent response format</li></ul><h3>DIAGRAM (ADD THIS SECTION)</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/553/1*v2UuRJtWgpfWnJHEWAaKrQ.png" /></figure><h3>Execution Modes</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/577/1*XsFQ9PvIXOOCj054t27QUQ.png" /></figure><h3>🔄 Example Flow</h3><p><strong>User Query:</strong><br> “Find related records for identifier X”</p><p><strong>Execution:</strong></p><ol><li>Agent detects:</li></ol><ul><li>Identifier-based input</li><li>Contextual intent</li></ul><p>2. Calls live system for base data</p><p>3. Queries search index for related records</p><p>4. Returns a unified response</p><h3>Benefits of Agentic Search Systems</h3><h4>✔️ Single Source of Truth</h4><p>Ensures data consistency</p><h4>✔️ Reduced Data Duplication</h4><p>Eliminates unnecessary synchronization</p><h4>✔️ Adaptive Query Handling</h4><p>Supports complex and mixed queries</p><h4>✔️ Improved User Experience</h4><p>Delivers more relevant results</p><h3>Challenges</h3><ul><li>Managing latency across systems</li><li>Designing robust routing logic</li><li>Handling hybrid query complexity</li><li>Maintaining consistent response formats</li></ul><h3>The Future of Enterprise Search</h3><p>Enterprise systems are evolving from:</p><blockquote><em>Static query pipelines</em></blockquote><p>to:</p><blockquote><em>Intelligent, decision-driven architectures</em></blockquote><blockquote>enables systems to:</blockquote><ul><li>Understand intent</li><li>Dynamically orchestrate data</li><li>Adapt to user needs</li></ul><h3>Conclusion</h3><p>Modern enterprise search requires more than indexing — it requires orchestration.</p><p>By integrating Agentic AI with systems like OpenSearch and live APIs, organizations can build:</p><ul><li>Scalable</li><li>Flexible</li><li>Future-ready platforms</li></ul><p>The future of search is not just about retrieving data — <br> it’s about retrieving the <strong>right data, from the right source, at the right time</strong>.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=f17dab9d1376" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Cracking the coding Interview —]]></title>
            <link>https://medium.com/@whoisinside/cracking-the-coding-interview-95ee5c3c6741?source=rss-58e13d89250f------2</link>
            <guid isPermaLink="false">https://medium.com/p/95ee5c3c6741</guid>
            <dc:creator><![CDATA[Lalsingh Ramavath]]></dc:creator>
            <pubDate>Mon, 21 Sep 2020 15:31:46 GMT</pubDate>
            <atom:updated>2020-09-21T15:31:46.054Z</atom:updated>
            <content:encoded><![CDATA[<h3>Longest Sub string with K Distinct Characters</h3><p>Lets discuss the problem : We need to find the longest subs string with no of</p><p>distinct characters.</p><pre>Input: String=&quot;araaci&quot;, K=1<br>Output: 2<br>Explanation: The longest substring with no more than &#39;1&#39; distinct characters is &quot;aa&quot;.</pre><pre>Input: String=&quot;cbbebi&quot;, K=3<br>Output: 5<br>Explanation: The longest substrings with no more than &#39;3&#39; distinct characters are &quot;cbbeb&quot; &amp; &quot;bbebi&quot;.</pre><p>a r a a c i</p><p>window start = a</p><p>window end = a and MaxLength = 0</p><p>Initially windowStart and WindowEnd point to the same character i.e (a) and at this state the max length would be ‘0’ . Lets shift the windowEnd to rite side until we reach the frequency of the character more than the ‘K’</p><p>Once we reach the frequency of the character more than to K , start shrinking the window by incrementing the windowStart and remove the shrink character from the Map . I will be taking the Hash Map here to store the string .</p><p>It would be consider as the medium problem .</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/930/1*30Uqu4AQ3SPnlvoQbzGksg.png" /></figure><h4>Time Complexity</h4><p><em>O</em>(<em>N</em>)</p><h4>Space Complexity</h4><p>The space complexity of the algorithm is O(K), as we will be storing a maximum of ‘K+1’ characters in the HashMap.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=95ee5c3c6741" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Cracking coding Interview — 2]]></title>
            <link>https://medium.com/@whoisinside/cracking-coding-interview-2-60f9fa7c680a?source=rss-58e13d89250f------2</link>
            <guid isPermaLink="false">https://medium.com/p/60f9fa7c680a</guid>
            <dc:creator><![CDATA[Lalsingh Ramavath]]></dc:creator>
            <pubDate>Sun, 20 Sep 2020 23:02:08 GMT</pubDate>
            <atom:updated>2020-09-20T23:02:08.624Z</atom:updated>
            <content:encoded><![CDATA[<p>Cracking coding Interview — 2</p><p><strong>Problem : Given an array of positive numbers and a positive number ‘k’, find the maximum sum of any contiguous sub array of size ‘k’.</strong></p><p>Lets think how we can solve the issue with best time and space complexity.</p><p>I/p : [2,4,5,3,6,7,8] , k =5 O/P = 29</p><p>Consider the first 5 elements will one one Window and Calculate the sum of the window i.e called WindowSum and initially it is ‘0’</p><p>Shift the window to next 5 elements i.e 4,5,3,6,7 and calculate the WindowSum . To get the current window sum subtract the first element ie. in array[0] .</p><p>windowSum = windowSum = array[windowStart]</p><p>Calculate the sum of sub array until we reaches the length of the array.</p><p>Lets write the simple Java program .</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/691/1*aejDQKZQQjwdNFIsLNqCOQ.png" /></figure><p>Time Complexity O(N)</p><p>Space-complexity O(1)</p><p>Please update the comment if you find the any best approach for this problem.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=60f9fa7c680a" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Cracking the coding interview Problems]]></title>
            <link>https://medium.com/@whoisinside/cracking-the-coding-interview-problems-afdc8ef8b6e5?source=rss-58e13d89250f------2</link>
            <guid isPermaLink="false">https://medium.com/p/afdc8ef8b6e5</guid>
            <dc:creator><![CDATA[Lalsingh Ramavath]]></dc:creator>
            <pubDate>Sun, 20 Sep 2020 21:36:21 GMT</pubDate>
            <atom:updated>2020-09-20T21:36:21.167Z</atom:updated>
            <content:encoded><![CDATA[<p>Cracking the coding interview Problems</p><ol><li>Given an array, find the average of all contiguous sub arrays of size ‘K’ in it</li></ol><p>In this article i just want to discuss the best approach to solve the problem and the time complexity .</p><p>Lets understand the problem .</p><p>i/p :Array: [5, 5, 2, 2, -1, 4, 1, 8, 2], K=5</p><p>We need to sum of the first 5 elements and calculate the average and store to new array .</p><p>5, 5, 2, 2, -1 = 13/5 = 2.6</p><p>5, 2, 2, -1, 4 = 12/5 = 2.4</p><p>2, 2, -1, 4, 1 = 8/5 = 1.6</p><p>2, -1, 4, 1, 8 = 14/5 = 2.8</p><p>-1, 4, 1, 8, 2 = 14/5 = 2.8</p><p>We can’t divide the array further because we need the 5 contiguous elements .</p><p>If we observer the problem we are taking the sun and avg of 5 elements and then shifting one element further and calculating the sum and avg of other 5 elements . If we assume 5 elements consider to me on Window then each iteration we are shifting window to rite side . Instead of taking two for loos and reading the value we are taking one loop and shifting the window so we are minimizing the space and time complexity .</p><p>Java Approach to solve the problem.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/876/1*v1IoTv3XL6VvTLLwjWb58g.png" /></figure><p>Every time we are reusing the previous sum and moving to next sum. When we are moving to next element and calculating the sum subtracting previous element (i.e .arr[windowStart].We are reusing the sum every time so the time complexity will be O(N). Please add the comment if you come across the other best approach .</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=afdc8ef8b6e5" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Angular Unit test Made easy ………]]></title>
            <link>https://medium.com/@whoisinside/angular-unit-test-made-easy-e11eed2c6457?source=rss-58e13d89250f------2</link>
            <guid isPermaLink="false">https://medium.com/p/e11eed2c6457</guid>
            <dc:creator><![CDATA[Lalsingh Ramavath]]></dc:creator>
            <pubDate>Wed, 16 Sep 2020 21:05:43 GMT</pubDate>
            <atom:updated>2020-09-16T21:05:43.186Z</atom:updated>
            <content:encoded><![CDATA[<p>Angular Unit test Made easy ………</p><h3>How To Unit Test Angular Component ,Service and HTML tags</h3><p>In this tutorial, you’ll see how to unit test Angular components, service and html input/Output tags or how to mock service dependency while testing an Angular component method.Sharing my own experiences while writing the angular unit tests some action based on the service success or error response.<br>In this tutorial we will be testing the service and associated components methods and its actions.</p><p>We will be wring the unit test in special file called .spec.ts file .We can put the no of spec files like for component we can create one spec file and for service we can create the other spec file .</p><h3>Angular Component</h3><p>To write the Unit test for the Angular component I will be using the Visual Code Studio and it has very nice cool features for build and test .</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1019/1*sqQ_dLEleIccVzKpCv1-EQ.png" /></figure><p>Here is the AppTestComponent.ts file .</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/427/1*Yw1NYx_EJj_koJ-JpGqhMg.png" /></figure><h3>Unit Test for Angular Component :</h3><p>Let’s start by writing unit test for testing method <em>getPostInfo .</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/636/1*4t_fMZTJZdHG6TnNSsxVmA.png" /></figure><p>In the beforeEach function for our test suite we <em>configure</em> a testing module using the TestBed class.</p><p>This creates a test <em>Angular Module</em> which we can use to instantiate components, perform dependency injection and so on.</p><p>We configure it in exactly the same way as we would configure a normal NgModule. If we have the navigation then add the Router module and if we have the form then add the forms module .</p><p>Test the HTML tags if we have any in .html file .</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/469/1*Jj8FnwIxlhP-Aax1nSwofA.png" /></figure><p>We can write positive test and negative test for each method in component and service too . Native Test means where we can expect the error and Positive test means we can get the expected result .</p><h3>Unit Test for Angular Service .</h3><p>Service which basically gets some Information.</p><p>Below is the service file:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/521/1*zEFwHt0Y5JN0ZvBwYZMU5w.png" /></figure><p>Here is the unit test for the service file .</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/661/1*YXwjsfQf_59Fc0ei_NoPIw.png" /></figure><p>Positive Unit Test .</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/629/1*dk2Zr9s12WELJ1MTlataTQ.png" /></figure><p>Negative Unit Test .</p><p>Test basically covered specific behavior when service fails loading data, the component redirects the user back .</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/622/1*lylfJandHQLqlHjcX3FafA.png" /></figure><p>Handle the tests with PUT, GET , POST etc .</p><p>Get :</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/572/1*S5IV17pxfTcykxU5VvXg0g.png" /></figure><p>POST :</p><p>While posting the data covered all the error codes like 400, 401 and 500 etc.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/813/1*UsyN5P4GamxEJACO0Xqr1Q.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*bWWazksOeJlt_k2WdliKnA.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/843/1*TcVgxra1X-65epKk7tqIsg.png" /></figure><p>Run the Tests :</p><p>ng test test-app</p><p>We can run the coverage report from command prompt also.</p><pre>ng test --code-coverage</pre><p>Karma.conf</p><p>coverageIstanbulReporter: {<br> dir: require(‘path’).join(__dirname, ‘../../coverage/testapp’),<br> reports: [‘html’, ‘lcovonly’],<br> fixWebpackSourcePaths: true<br> },</p><p>In case of any questions i will answer through the comments and don’t forget to Clap if you like it .</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e11eed2c6457" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Machine Learning Algorithms]]></title>
            <link>https://medium.com/@whoisinside/machine-learning-algorithms-c9076981b3e9?source=rss-58e13d89250f------2</link>
            <guid isPermaLink="false">https://medium.com/p/c9076981b3e9</guid>
            <dc:creator><![CDATA[Lalsingh Ramavath]]></dc:creator>
            <pubDate>Fri, 04 Sep 2020 04:18:22 GMT</pubDate>
            <atom:updated>2020-09-04T04:18:22.109Z</atom:updated>
            <content:encoded><![CDATA[<h3>Case Study (Term Deposit Sale) Using Logistic Regression and ensemble models (Bagging and Boosting)</h3><p>Goal Using the data collected from existing customers, build a model that will help the marketing team identify potential customers who are relatively more likely to subscribe term deposit and thus increase their hit ratio.</p><p>Resources Available The historical data for this project is available in file <a href="https://archive.ics.uci.edu/ml/datasets/Bank+Marketing">https://archive.ics.uci.edu/ml/datasets/Bank+Marketing</a></p><p>Exploratory data Analysis:</p><p><strong>Attribute Information:</strong></p><p>Input variables:<br># bank client data:<br>1 — age (numeric)<br>2 — job : type of job (categorical: ‘admin.’,’blue-collar’,’entrepreneur’,’housemaid’,’management’,’retired’,’self-employed’,’services’,’student’,’technician’,’unemployed’,’unknown’)<br>3 — marital : marital status (categorical: ‘divorced’,’married’,’single’,’unknown’; note: ‘divorced’ means divorced or widowed)<br>4 — education (categorical: ‘basic.4y’,’basic.6y’,’basic.9y’,’high.school’,’illiterate’,’professional.course’,’university.degree’,’unknown’)<br>5 — default: has credit in default? (categorical: ‘no’,’yes’,’unknown’)<br>6 — housing: has housing loan? (categorical: ‘no’,’yes’,’unknown’)<br>7 — loan: has personal loan? (categorical: ‘no’,’yes’,’unknown’)<br># related with the last contact of the current campaign:<br>8 — contact: contact communication type (categorical: ‘cellular’,’telephone’)<br>9 — month: last contact month of year (categorical: ‘jan’, ‘feb’, ‘mar’, …, ‘nov’, ‘dec’)<br>10 — day_of_week: last contact day of the week (categorical: ‘mon’,’tue’,’wed’,’thu’,’fri’)<br>11 — duration: last contact duration, in seconds (numeric). Important note: this attribute highly affects the output target (e.g., if duration=0 then y=’no’). Yet, the duration is not known before a call is performed. Also, after the end of the call y is obviously known. Thus, this input should only be included for benchmark purposes and should be discarded if the intention is to have a realistic predictive model.<br># other attributes:<br>12 — campaign: number of contacts performed during this campaign and for this client (numeric, includes last contact)<br>13 — pdays: number of days that passed by after the client was last contacted from a previous campaign (numeric; 999 means client was not previously contacted)<br>14 — previous: number of contacts performed before this campaign and for this client (numeric)<br>15 — poutcome: outcome of the previous marketing campaign (categorical: ‘failure’,’nonexistent’,’success’)<br># social and economic context attributes<br>16 — emp.var.rate: employment variation rate — quarterly indicator (numeric)<br>17 — cons.price.idx: consumer price index — monthly indicator (numeric)<br>18 — cons.conf.idx: consumer confidence index — monthly indicator (numeric)<br>19 — euribor3m: euribor 3 month rate — daily indicator (numeric)<br>20 — nr.employed: number of employees — quarterly indicator (numeric</p><p>Import all the necessary libraries:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/573/1*BN9qtRK9DoDhreHZ3wVThg.png" /></figure><p>Loading the data into Data frame</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/423/1*Ty8yy23e5kHSoaWa5VbRDw.png" /></figure><p>View the data tail and head from the Data frame:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*XzGNbJXHhB8Jt82niz3apQ.png" /></figure><p>View the describe and Info:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*keTY0oyXWRtWjUPX77wTzQ.png" /></figure><p>Check the transpose and shape to see no of columns , Mean , Count and std.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/759/1*msP6WPXbstw5mTbyJnkDOA.png" /></figure><p>It is very Informant to check the missing values while preparing the data .</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*qAqtTvpu3L7HmbU3mO87fQ.png" /></figure><pre><em>#Univariate  Analysis :Analysis done based only on one variable<br></em>sns.FacetGrid(bank_data , hue = &quot;Target&quot; , size =5).map(sns.distplot,&quot;age&quot;).add_legend()<br></pre><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*3knGSVK1XGNt9fyc8sbsuA.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/630/1*csn5NaHMD4Qmu_UOkh6Xhw.png" /></figure><pre>sns.FacetGrid(bank_data , hue = &quot;Target&quot; , size = 5).map(sns.countplot,&quot;loan&quot;).add_legend()</pre><figure><img alt="" src="https://cdn-images-1.medium.com/max/691/1*dolP7HtUanyP8wp3tLB_lg.png" /></figure><pre><em>#Bivariate analysis:</em><br><em>#relation ship between the two variables</em><br><em>#boxplot</em><br>sns.boxplot(x=&#39;Target&#39;, y= &#39;age&#39;, data = bank_data )</pre><figure><img alt="" src="https://cdn-images-1.medium.com/max/684/1*y1GLNEPCioK86TKLIPB4hQ.png" /></figure><pre>sns.violinplot(x=&#39;Target&#39;, y= &#39;duration&#39;, data = bank_data )</pre><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*RYGATrLNbEoEz50fKMBWZA.png" /></figure><p>#Multivariate analysis: Multiple Variables</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*s3yTadG00rZ5LCOWeARsUw.png" /></figure><p>bank_data.corr() # displays the correlation between every possible pair of attributes as a dataframe</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*FhthqqNzSV6BTAHM4iBMFA.png" /></figure><pre>#heat Map and scatterplot<br>sns.heatmap(bank_data.corr(), annot=True)  # plot the correlation coefficients as a heatmap</pre><figure><img alt="" src="https://cdn-images-1.medium.com/max/956/1*XiSYvr47hj-xmWnEM5sOAw.png" /></figure><p>#Prepare the data for analytics and remove the duplicated<br>bank_data = bank_data.drop_duplicates()</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/444/1*luSLJnGJiWl-J-EA69WzGw.png" /></figure><p>Replace the above objects with numerical data so data easily can be converted training set and test set.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*4heHhx4mAAKUt-3U21pzVg.png" /></figure><p>Complete data set with all the numerical values.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*3aZfbq2DKTTUU47XnT26gA.png" /></figure><p>Drop the columns in case any column is not useful.</p><p>Split the data into 70 : 30 training and test and make sure take the target variable into training and test data . Here “ Target” is the target variable.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*EuxRxHsnWxNkZlXVlLWmMg.png" /></figure><p>LogisticRegression and predict the model score :</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*v-wN55o56rrmi3DWYrS5jA.png" /></figure><p>Confusion Matrix:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*t7RV3Z-wk6vJTJzMrRBjIw.png" /></figure><p>Calculate Accuracy ,Precision,Recall and F1 score:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*jQgm3r5s16pDz6gDtMBbCA.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/693/1*_HAnotrcSWMfrGzFuX-jaw.png" /></figure><p>Ensemble methods:</p><p>#Building the Decision Tree , Baggind an<br>#Build Decision Tree Model</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*1J0TxM0N6MahXc0WDIlKgg.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*-ikRZkhizGdcyEW21SN2cw.png" /></figure><p>Confusion Matrix for DecisionTree:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*tt_u-wZkNse17BRH6q8brA.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/812/1*OqtCC_nTOzPbacvFJ_BNdA.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*d6vyIsYGRHQu0zScni6t1w.png" /></figure><p>Accuracy for Decision Tree and Random Forest:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*t9Tr44kr-DNZ1GLjO0WyAg.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*j5MHk-_T5MOIt8S_jjcZyQ.png" /></figure><p>Bagging :</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*usZYOhQAZVVvb-Uinuen1g.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/981/1*5tlSq2donrMwoGSCC3rA1Q.png" /></figure><p>GradientBoosting:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*E4A5YGUZuRP-HsLblYzn-w.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*EyJ2wfYMCzLZ4_9J-wmdiQ.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/987/1*vSRJrawz0GKa7DVw8zrUcQ.png" /></figure><p>We have predicted the different models and predicted the accuracy score using the Ensemble Techniques.</p><ul><li>***** If you Like the Case Study Please Clap Clap Clap ******</li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=c9076981b3e9" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>