<?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 Ameya Thorat on Medium]]></title>
        <description><![CDATA[Stories by Ameya Thorat on Medium]]></description>
        <link>https://medium.com/@ameyathorat?source=rss-3923c775a9ba------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*9k21gbWgG6C2AdAHHXfVQw@2x.jpeg</url>
            <title>Stories by Ameya Thorat on Medium</title>
            <link>https://medium.com/@ameyathorat?source=rss-3923c775a9ba------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Mon, 16 Nov 2026 18:07:49 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@ameyathorat/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[Machine Learning Resources — MIT course]]></title>
            <link>https://ameyathorat.medium.com/machine-learning-resources-mit-course-96d6c138e2bb?source=rss-3923c775a9ba------2</link>
            <guid isPermaLink="false">https://medium.com/p/96d6c138e2bb</guid>
            <category><![CDATA[machine-learning]]></category>
            <dc:creator><![CDATA[Ameya Thorat]]></dc:creator>
            <pubDate>Sat, 24 Feb 2024 18:23:08 GMT</pubDate>
            <atom:updated>2024-02-24T18:23:08.968Z</atom:updated>
            <content:encoded><![CDATA[<h3>Machine Learning Resources — MIT course</h3><p>From Miruna Clinciu, Teaching Assistant, MIT, Designing &amp; Building AI Products and Services, Feb 23, 2024</p><p>Miruna Clinciu (She/Her) , Ken Rhines Designing and Building AI Products and Services February 22, 2024 at 8:02am</p><p>I hope this message finds you well.</p><p>Hi there,</p><p>In response to numerous inquiries regarding additional learning materials, I have compiled a list of resources for those who wish to delve deeper into the realm of machine learning.</p><p>It’s important to note that while these materials can offer valuable insights, the comprehensive information provided in the course videos and content should be more than sufficient for a thorough understanding</p><p>==========</p><h3>Machine Learning — short description + resources</h3><h3>1. Gradient descent</h3><p>Gradient Descent is an algorithm that solves optimization problems using first-order iterations. Since it is designed to find the local minimum of a differential function, gradient descent is widely used in machine learning models to find the best parameters that minimize the model’s cost function.</p><p>Let’s say you are playing a game where the players are at the top of a mountain, and they are asked to reach the lowest point of the mountain. Additionally, they are blindfolded. So, what approach do you think would make you reach the lake?</p><p>Tthe sloppiness of the hill is high at the top, one single slip, and you are done! The sloppiness when you reach the bottom of the hill is minimal or close to or equal to zero, the same scenario works in Mathematics as well.</p><h3>Resources:</h3><p>What Does Gradient Descent Actually Mean</p><p><a href="https://www.analyticsvidhya.com/blog/2020/10/what-does-gradient-descent-actually-mean/Links">https://www.analyticsvidhya.com/blog/2020/10/what-does-gradient-descent-actually-mean/Links</a> to an external site. Gradient Descent Algorithm and Its Variants</p><p><a href="https://towardsdatascience.com/gradient-descent-algorithm-and-its-variants-10f652806a3Links">https://towardsdatascience.com/gradient-descent-algorithm-and-its-variants-10f652806a3Links</a> to an external site. Optimization Methods GD</p><p><a href="https://github.com/y33-j3T/Coursera-Deep-Learning/blob/master/Improving%20Deep%20Neural%20Networks%20Hyperparameter%20tuning,%20Regularization%20and%20Optimization/week6/Optimization_methods_v1b.ipynbLinks">https://github.com/y33-j3T/Coursera-Deep-Learning/blob/master/Improving%20Deep%20Neural%20Networks%20Hyperparameter%20tuning,%20Regularization%20and%20Optimization/week6/Optimization_methods_v1b.ipynbLinks</a> to an external site. <a href="https://medium.com/@jackyfeng530/machine-learning-gradient-descent-8681c40d27c3Links">https://medium.com/@jackyfeng530/machine-learning-gradient-descent-8681c40d27c3Links</a> to an external site.</p><h3>2. Learning Rate</h3><ul><li>the learning rate is a tuning parameter in an optimization algorithm that determines the step size at each iteration while moving toward a minimum of a loss function</li></ul><p>If the learning rate is too high then it bounces to convex function but not reached local minimum efficiently</p><p>If the learning rate is too small then it reached to local minimum but it takes too much time</p><p>It must be chosen carefully to end up with local minima.</p><p>If the learning rate is too high, we might OVERSHOOT the minima and keep bouncing, without reaching the minima If the learning rate is too small, the training might turn out to be too long a) Learning rate is optimal, model converges to the minimum</p><p>b) Learning rate is too small, it takes more time but converges to the minimum</p><p>c) Learning rate is higher than the optimal value, it overshoots but converges ( 1/C &lt; η &lt;2/C)</p><p>d) Learning rate is very large, it overshoots and diverges, moves away from the minima, performance decreases on learning</p><h3>Resources:</h3><p>Understand the Impact of Learning Rate on Neural Network Performance</p><p><a href="https://machinelearningmastery.com/understand-the-dynamics-of-learning-rate-on-deep-learning-neural-networks/Links">https://machinelearningmastery.com/understand-the-dynamics-of-learning-rate-on-deep-learning-neural-networks/Links</a> to an external site.</p><h3>3. Batch-Gradient Descent (Vanilla)</h3><p>It is also called as vanilla gradient descent technique. In this technique we calculates the error at each observation but perform the update only when all observation done evaluating</p><p>Batch gradient descent technique is computationally expensive technique. It is very slow and large data-sets does not load into memory.</p><h3>4. Stochastic Gradient Descent (SGD)</h3><p>SGD performs the weight updates on each observation with every iterations. here we first need to shuffle the data-set so that we get completely randomised data set.</p><p>As our data-set is randomised data-set and error and weight updated for each example gradient weights are arriving at global minima instead of stuck in local minima.</p><p>SGD is faster than batch gradient technique. This technique also computationally redundant because it computes each example at a time</p><p>SGD is noisy — Due to frequent updates, the steps taken towards the minima are very noisy. This can often lean the gradient descent into other directions. Also, due to noisy steps, it may take longer to achieve convergence to the minima of the loss function.</p><h3>5. Mini-Batch Gradient Descent</h3><p>This technique is a combination of sgd and batch gradient technique. First it split the data-set into small batches and then perform updates on each batches. Here as we take data in batch samples it removes the noise which is variance of the weight update. It is speedy technique as compared to both batch and sgd method.</p><h3>Resources:</h3><p>Gradient descent: Why and How ?</p><p><a href="https://medium.com/analytics-vidhya/gradient-descent-why-and-how-e369950ae7d3#:~:text=Mini%2DBatch%20Gradient%20Descent%20%3A%2D,perform%20updates%20on%20each%20batches">https://medium.com/analytics-vidhya/gradient-descent-why-and-how-e369950ae7d3#:~:text=Mini%2DBatch%20Gradient%20Descent%20%3A%2D,perform%20updates%20on%20each%20batches</a> Links to an external site..</p><h3>6. Cost function</h3><p>In the case of gradient descent, the objective is to find a line of best fit for some given inputs, or X values, and any number of Y values, or outputs. A cost function is defined as:</p><p>Cost function — function that maps an event or values of one or more variables onto a real number intuitively representing some “cost” associated with the event.</p><h3>Resources:</h3><p><a href="https://medium.com/@lachlanmiller_52885/understanding-and-calculating-the-cost-function-for-linear-regression-39b8a3519fcbLinks">https://medium.com/@lachlanmiller_52885/understanding-and-calculating-the-cost-function-for-linear-regression-39b8a3519fcbLinks</a> to an external site.</p><h3>7. Transformer</h3><p>A transformer is a deep learning Links to an external site.model that adopts the mechanism of self-attention Links to an external site., differentially weighting the significance of each part of the input data.</p><h3>Resources:</h3><p><a href="https://www.theaidream.com/post/transformer-neural-network-in-deep-learning-explained">https://www.theaidream.com/post/transformer-neural-network-in-deep-learning-explained</a></p><h3>8. Self-attention</h3><p>A self-attention module works by comparing every word in the sentence to every other word in the sentence, including itself, and reweighing the word embeddings of each word to include contextual relevance. It takes in n word embeddings without context and returns n word embeddings with contextual information.</p><h3>Resources:</h3><p><a href="https://towardsdatascience.com/an-intuitive-explanation-of-self-attention-4f72709638e1Links">https://towardsdatascience.com/an-intuitive-explanation-of-self-attention-4f72709638e1Links</a> to an external site.</p><h3>9 &amp; 10. Regularization L1/L2</h3><p>Overfitting is a phenomenon that occurs when a machine learning or statistics model is tailored to a particular dataset and is unable to generalise to other datasets. This usually happens in complex models, like deep neural networks.</p><p>Regularisation is a process of introducing additional information in order to prevent overfitting.</p><p>A linear regression model that implements L1 norm for regularisation is called lasso regression, and one that implements (squared) L2 norm for regularisation is called ridge regression.</p><p>L2 — combats overfitting by forcing weights to be small, but not making them exactly 0. L1 regularization penalizes the sum of absolute values of the weights, whereas L2 regularization penalizes the sum of squares of the weights. The L1 regularization solution is sparse. The L2 regularization solution is non-sparse. L2 regularization doesn’t perform feature selection, since weights are only reduced to values near 0 instead of 0. L1 regularization has built-in feature selection. L1 regularization is robust to outliers, L2 regularization is not. L1 regularization gives output in binary weights from 0 to 1 for the model’s features and is adopted for decreasing the number of features in a huge dimensional dataset. L2 regularization disperse the error terms in all the weights that leads to more accurate customized final models. The main intuitive difference between the L1 and L2 regularization is that L1 regularization tries to estimate the median of the data while the L2 regularization tries to estimate the mean of the data to avoid overfitting.</p><h3>11. Difference between Discriminative and Generative Models</h3><p>Discriminative models draw boundaries in the data space, while generative models try to model how data is placed throughout the space. A generative model focuses on explaining how the data was generated, while a discriminative model focuses on predicting the labels of the data.</p><p>Discriminative models:</p><p>Logistic regression Scalar Vector Machine (SVMs) ‌Traditional neural networks ‌Nearest neighbor Conditional Random Fields (CRFs) Decision Trees and Random Forest Generative Models:</p><p>‌Naïve Bayes Bayesian networks Markov random fields ‌Hidden Markov Models (HMMs) Latent Dirichlet Allocation (LDA) Generative Adversarial Networks (GANs)</p><h3>Resources:</h3><p>2023’s Best Guide to Descriptive &amp; Generative Machine Learning Models</p><p><a href="https://www.analyticsvidhya.com/blog/2021/07/deep-understanding-of-discriminative-and-generative-models-in-machine-learning/Links">https://www.analyticsvidhya.com/blog/2021/07/deep-understanding-of-discriminative-and-generative-models-in-machine-learning/Links</a> to an external site.</p><h3>12. Logistic Regression</h3><p>Logistic regression is applied to predict the categorical dependent variable. In other words, it’s used when the prediction is categorical, for example, yes or no, true or false, 0 or 1.</p><h3>Resources:</h3><p>Conceptual Understanding of Logistic Regression for Data Science Beginners</p><p><a href="https://www.analyticsvidhya.com/blog/2021/08/conceptual-understanding-of-logistic-regression-for-data-science-beginners/Links">https://www.analyticsvidhya.com/blog/2021/08/conceptual-understanding-of-logistic-regression-for-data-science-beginners/Links</a> to an external site.</p><h3>13. Bias — Variance Tradeoff</h3><p>Bias is the simplifying assumptions made by the model to make the target function easier to approximate. Variance is the amount that the estimate of the target function will change given different training data.</p><p>“Bias and variance are complements of each other” The increase of one will result in the decrease of the other and vice versa.</p><h3>Resources:</h3><p>Gentle Introduction to the Bias-Variance Trade-Off in Machine Learning</p><p><a href="https://machinelearningmastery.com/gentle-introduction-to-the-bias-variance-trade-off-in-machine-learning/Links">https://machinelearningmastery.com/gentle-introduction-to-the-bias-variance-trade-off-in-machine-learning/Links</a> to an external site.</p><h3>14. Normalization</h3><p>The goal of normalization is to transform features to be on a similar scale.</p><h3>15. Gradient boosting</h3><p>Gradient boosting is a machine learning boosting type. It strongly relies on the prediction that the next model will reduce prediction errors when blended with previous ones.</p><h3>16. Naive-Bayes</h3><p>A Naive Bayes classifier is a probabilistic machine learning model that’s used for classification task.</p><p>Based on the Bayes theorem.</p><p>Bayes Theorem:</p><p>Using Bayes theorem, we can find the probability of A happening, given that B has occurred. Here, B is the evidence and A is the hypothesis. The assumption made here is that the predictors/features are independent. That is presence of one particular feature does not affect the other. Hence it is called naive.</p><h3>Resources:</h3><p>What’s ‘naive’ about Naive Bayes Classifier?</p><p><a href="https://pub.towardsai.net/whats-naive-about-naive-bayes-classifier-dab83d98dfe0Links">https://pub.towardsai.net/whats-naive-about-naive-bayes-classifier-dab83d98dfe0Links</a> to an external site.</p><h3>17. Linear Regression</h3><p>Linear regression Links to an external site.is an analysis that assesses whether one or more predictor variables explain the dependent (criterion) variable. The regression has five key assumptions:</p><p>Linear relationship (There must be a linear relationship between the outcome variable and the independent variables. Scatterplots can show whether there is a linear or curvilinear relationship.)</p><p>Multivariate normality (Multivariate Normality–Multiple regression assumes that the residuals are normally distributed.)</p><p>No or little multicollinearity Links to an external site.(Multicollinearity occurs when two or more independent variables are highly correlated with one another in a regression model. )</p><p>No auto-correlation (No autocorrelation refers to a situation in which no identifiable relationship exists between the values of the error term.)</p><p>Homoscedasticity Links to an external site.(This assumption states that the variance of error terms are similar across the values of the independent variables. A plot of standardized residuals versus predicted values can show whether points are equally distributed across all values of the independent variables.)</p><h3>Resources:</h3><p>Patterns of Autocorrelation</p><p><a href="https://www.dummies.com/article/business-careers-money/business/economics/patterns-of-autocorrelation-156434/Links">https://www.dummies.com/article/business-careers-money/business/economics/patterns-of-autocorrelation-156434/Links</a> to an external site.</p><h3>18. Decision Tree</h3><p>A decision tree is a non-parametric supervised learning algorithm, which is utilized for both classification and regression tasks. It has a hierarchical, tree structure, which consists of a root node, branches, internal nodes and leaf nodes.</p><p>A decision tree starts with a root node, which does not have any incoming branches. The outgoing branches from the root node then feed into the internal nodes, also known as decision nodes. Based on the available features, both node types conduct evaluations to form homogenous subsets, which are denoted by leaf nodes, or terminal nodes. The leaf nodes represent all the possible outcomes within the dataset.</p><h3>Resources:</h3><p>What is a Decision Tree?</p><p><a href="https://www.ibm.com/uk-en/topics/decision-treesLinks">https://www.ibm.com/uk-en/topics/decision-treesLinks</a> to an external site.</p><h3>19. K-means</h3><p>The working of the K-Means algorithm is explained in the below steps:</p><p>Step-1: Select the number K to decide the number of clusters.</p><p>Step-2: Select random K points or centroids. (It can be other from the input dataset).</p><p>Step-3: Assign each data point to their closest centroid, which will form the predefined K clusters.</p><p>Step-4: Calculate the variance and place a new centroid of each cluster.</p><p>Step-5: Repeat the third steps, which means reassign each datapoint to the new closest centroid of each cluster.</p><p>Step-6: If any reassignment occurs, then go to step-4 else go to FINISH.</p><p>Step-7: The model is ready.</p><h3>Elbow Method</h3><p>It executes the K-means clustering on a given dataset for different K values (ranges from 1–10). For each value of K, calculates the WCSS value. Plots a curve between calculated WCSS values and the number of clusters K. The sharp point of bend or a point of the plot looks like an arm, then that point is considered as the best value of K.</p><h3>20. Random Forest Importance</h3><p>Random Forests is a kind of a Bagging Algorithm that aggregates a specified number of decision trees. The tree-based strategies used by random forests naturally rank by how well they improve the purity of the node, or in other words a decrease in the impurity (Gini impurity) over all trees. Nodes with the greatest decrease in impurity happen at the start of the trees, while notes with the least decrease in impurity occur at the end of trees. Thus, by pruning trees below a particular node, we can create a subset of the most important feature</p><h3>21. Principal Components Analysis (PCA)</h3><p>Standardize the range of continuous initial variables Compute the covariance matrix to identify correlations Compute the eigenvectors and eigenvalues of the covariance matrix to identify the principal components Create a feature vector to decide which principal components to keep Recast the data along the principal components axes</p><h3>22. SVM</h3><p><a href="https://web.mit.edu/6.034/wwwbob/svm-notes-long-08.pdfLinks">https://web.mit.edu/6.034/wwwbob/svm-notes-long-08.pdfLinks</a> to an external site. <a href="https://www-users.cse.umn.edu/~kumar001/dmbook/slides/chap4_svm.pdfLinks">https://www-users.cse.umn.edu/~kumar001/dmbook/slides/chap4_svm.pdfLinks</a> to an external site. <a href="https://web.mit.edu/6.034/wwwbob/svm-notes-long-08.pdfLinks">https://web.mit.edu/6.034/wwwbob/svm-notes-long-08.pdfLinks</a> to an external site. <a href="https://phillipi.github.io/6.882/2020/notes/6.036_notes.pdfLinks">https://phillipi.github.io/6.882/2020/notes/6.036_notes.pdfLinks</a> to an external site. <a href="https://www.youtube.com/watch?v=gUzEN2TxnxELinks">https://www.youtube.com/watch?v=gUzEN2TxnxELinks</a> to an external site.</p><h3>BOOKS and Courses:</h3><ul><li>Introduction to Data Mining by Michael Steinbach, Pang-Ning Tan, and Vipin Kumar Link: <a href="https://www-users.cse.umn.edu/~kumar001/dmbook/">https://www-users.cse.umn.edu/~kumar001/dmbook/</a></li><li>Artificial Intelligence: A Modern Approach 3rd + 4th versions Link: <a href="https://github.com/yanshengjia/ml-road/blob/master/resources/Artificial%20Intelligence%20-%20A%20Modern%20Approach%20(3rd%20Edition).pdf">https://github.com/yanshengjia/ml-road/blob/master/resources/Artificial%20Intelligence%20-%20A%20Modern%20Approach%20(3rd%20Edition).pdf</a></li><li>Pattern Recognition and Machine Learning by Christopher Bishop Link: <a href="https://www.microsoft.com/en-us/research/uploads/prod/2006/01/Bishop-Pattern-Recognition-and-Machine-Learning-2006.pdf">https://www.microsoft.com/en-us/research/uploads/prod/2006/01/Bishop-Pattern-Recognition-and-Machine-Learning-2006.pdf</a></li><li>Deep Learning by Ian Goodfellow and Yoshua Bengio and Aaron Courville Link <a href="https://www.deeplearningbook.org/">https://www.deeplearningbook.org/</a></li></ul><p>LSTM <a href="http://colah.github.io/posts/2015-08-Understanding-LSTMs/">http://colah.github.io/posts/2015-08-Understanding-LSTMs/</a> CNN <a href="https://e2eml.school/how_convolutional_neural_networks_work.html">https://e2eml.school/how_convolutional_neural_networks_work.html</a> Math for Gradients: <a href="https://medium.com/swlh/understand-the-math-for-neural-networks-ae4f75c7ccd9">https://medium.com/swlh/understand-the-math-for-neural-networks-ae4f75c7ccd9</a> Math DL: <a href="https://explained.ai/matrix-calculus/">https://explained.ai/matrix-calculus/</a></p><p>A/B testing: <a href="https://medium.com/@talraviv/please-please-dont-a-b-test-that-980a9630e4fb">https://medium.com/@talraviv/please-please-dont-a-b-test-that-980a9630e4fb</a></p><p>Data drift vs Model drift <a href="https://deepchecks.com/data-drift-vs-concept-drift-what-are-the-main-differences/">https://deepchecks.com/data-drift-vs-concept-drift-what-are-the-main-differences/</a> <a href="https://www.datacamp.com/tutorial/understanding-data-drift-model-drift">https://www.datacamp.com/tutorial/understanding-data-drift-model-drift</a></p><p>Concept drift <a href="https://www.coursera.org/lecture/gcp-production-ml-systems/concept-drift-kcTDO?utm_source=link&amp;utm_medium=page_share&amp;utm_content=vlp&amp;utm_campaign=top_button">https://www.coursera.org/lecture/gcp-production-ml-systems/concept-drift-kcTDO?utm_source=link&amp;utm_medium=page_share&amp;utm_content=vlp&amp;utm_campaign=top_button</a></p><p>Understanding Neural Networks =&gt;</p><p>Network Depth: Depending on the problem, shallow networks might be sufficient. We consider deep neural networks when we have more than 3 layes. However, we have problems associated with deep neural networks: vanishing or exploding gradients</p><p>Number of neurons: Too few neurons may cause underfitting, while too many can lead to overfitting. <a href="https://machinelearningmastery.com/introduction-to-regularization-to-reduce-overfitting-and-improve-generalization-error/">https://machinelearningmastery.com/introduction-to-regularization-to-reduce-overfitting-and-improve-generalization-error/</a> Check this thread: <a href="https://www.researchgate.net/post/why_the_prediction_or_the_output_of_neural_network_does_not_change_during_the_test_phase">https://www.researchgate.net/post/why_the_prediction_or_the_output_of_neural_network_does_not_change_during_the_test_phase</a> <a href="https://www.linkedin.com/pulse/choosing-number-hidden-layers-neurons-neural-networks-sachdev">https://www.linkedin.com/pulse/choosing-number-hidden-layers-neurons-neural-networks-sachdev</a></p><p>Other resources: <a href="http://neuralnetworksanddeeplearning.com/">http://neuralnetworksanddeeplearning.com/</a> <a href="https://tanthiamhuat.files.wordpress.com/2018/03/deeplearningwithpython.pdf">https://tanthiamhuat.files.wordpress.com/2018/03/deeplearningwithpython.pdf</a> <a href="https://machinelearningmastery.com/deep-learning-with-python/">https://machinelearningmastery.com/deep-learning-with-python/</a> <a href="http://silverio.net.br/heitor/disciplinas/eeica/papers/Livros/%5BChollet%5D-Deep_Learning_with_Python.pdf">http://silverio.net.br/heitor/disciplinas/eeica/papers/Livros/[Chollet]-Deep_Learning_with_Python.pdf</a> <a href="https://news.mit.edu/2017/explained-neural-networks-deep-learning-0414">https://news.mit.edu/2017/explained-neural-networks-deep-learning-0414</a> <a href="https://www.deeplearningbook.org/">https://www.deeplearningbook.org/</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=96d6c138e2bb" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Preparing for a Product Manager Interview — webinar]]></title>
            <link>https://ameyathorat.medium.com/preparing-for-a-product-manager-interview-webinar-cdd2330f9f05?source=rss-3923c775a9ba------2</link>
            <guid isPermaLink="false">https://medium.com/p/cdd2330f9f05</guid>
            <category><![CDATA[faang-interview]]></category>
            <category><![CDATA[interview]]></category>
            <category><![CDATA[faang]]></category>
            <category><![CDATA[interview-questions]]></category>
            <category><![CDATA[product-management]]></category>
            <dc:creator><![CDATA[Ameya Thorat]]></dc:creator>
            <pubDate>Thu, 13 May 2021 19:50:29 GMT</pubDate>
            <atom:updated>2021-05-13T20:02:10.633Z</atom:updated>
            <content:encoded><![CDATA[<h3>Preparing for a Product Manager Interview — webinar</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*N-K8Y-39ZM-sOSnUReix8g.jpeg" /><figcaption>Photo by <a href="https://unsplash.com/@steve3p_0?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Steve Halama</a> on <a href="https://unsplash.com/s/photos/interview?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Unsplash</a></figcaption></figure><p>Here is a link to the a webinar I recently conducted for Product Management. Covers a few tips on how to answer behavioral questions in product management interviews.</p><p>Link to the slides: <a href="https://www.slideshare.net/secret/1w6FdFETZTghGe">https://www.slideshare.net/secret/1w6FdFETZTghGe</a></p><p><a href="https://www.linkedin.com/video/live/urn:li:ugcPost:6797262727938088961/">Product School on LinkedIn: #ProductManagement | 589 comments</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=cdd2330f9f05" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[TLDR the question that is asked in almost every interview, tips on do’s and don’ts for this…]]></title>
            <link>https://ameyathorat.medium.com/why-tell-me-about-yourself-is-an-important-question-for-product-management-interviews-74132ccee50b?source=rss-3923c775a9ba------2</link>
            <guid isPermaLink="false">https://medium.com/p/74132ccee50b</guid>
            <category><![CDATA[interview]]></category>
            <category><![CDATA[interview-questions]]></category>
            <category><![CDATA[product-management]]></category>
            <category><![CDATA[interview-tips]]></category>
            <category><![CDATA[product-manager-interview]]></category>
            <dc:creator><![CDATA[Ameya Thorat]]></dc:creator>
            <pubDate>Thu, 06 May 2021 17:55:45 GMT</pubDate>
            <atom:updated>2024-02-24T18:21:26.535Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/800/1*GOKzBGpLaNTRHIVm3_r-mw.jpeg" /><figcaption>Photo by <a href="https://unsplash.com/@jexo?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Jexo</a> “Tell me about yourself” is an important question for Product Management Interviews?</figcaption></figure><p>TLDR the question that is asked in almost every interview, tips on do’s and don’ts for this question.</p><p>Before I start, here is a quick summary of my background. I have been a product manager for the last 10+ years in the consumer as well as enterprise space (mostly consumer). Most recently, I was a Product Leader at Amazon (for the last 4.5 years) and worked at various teams (Alexa, Amazon Identity, Amazon Business). Currently, I am a Group Product Manager at <a href="https://finance.yahoo.com/quote/OKTA?.tsrc=applewf">Okta</a> on the Access Management Team.</p><p>In my product career, I have conducted hundreds of interviews and only a handful of times I have come across candidates who have not recited their resumes when I asked “Tell me about yourself”. Most candidates take this question for granted and will answer this question by either casually mentioning their past work-ex or just do a read through of their resume. They seem to miss the nuance behind the question. People are not really interested in your resume because (guess what) if they had doubts on your accomplishment or candidacy, then you wouldn’t have gotten the call in the first place. Interviewers want to know more about YOU which is not stated in the resume , traits such as what drives you? what kind of a leader are you? what kind of a work environment do you thrive in? … you get the point.</p><p><strong>Why is this an important question?</strong></p><p>The short answer is, this question is asked in every single interview and often this is the first question that you will face. In other words, this is your first chance to create an impression. How you answer this can set you apart from the rest and make your candidacy a memorable experience for the interviewer.</p><p>let’s take an example. Here is a common response I have seen, when asked to tell me about yourself:</p><p>Most Common: <em>Hello , my name is ABC, currently I work at company A as a product manager for &lt;product area&gt; and I own &lt;1 or 2 areas of responsibility&gt;. Prior to this I was at company B (followed by similar narrative as for company A). I have an undergrad in CS and an MBA in marketing.</em></p><p>There may be some variations/ additions to the above narrative but more of less I have seen a similar structure from candidates. Do you think the above example tell the interviewer anything that your resume did not? The interviewer could have might as well read your resume (again) and derived more information than asking you the question.</p><p><strong>How could you answer it?</strong></p><p>Remember you may have at most 60–90 seconds for this so keep it crisp. I will not provide a verbatim answer, but rather mention a few key points that you should include in your response. here is a rough structure,</p><ol><li>One line summary of your career thus far, “I have N years of PM experience in the enterprise space”</li><li>Talk about your educational background (especially if it aligns with the role requirement) e.g. undergraduate degree in computer science and MBA in marketing</li><li>Your current company, position, tenure and an achievement in that role (state metric)</li><li>One to two lines on previous work experience.</li><li>What motivates you or environment that you thrive in (fast paced, process oriented etc.)</li></ol><p>In summary, you will need to take time and instrospect to answer this question thoughtfully and convey to the interviewer from the very first sentence in the interview why and how you are better than other candidates.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=74132ccee50b" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Why do people tip generously (mostly in America)?]]></title>
            <link>https://ameyathorat.medium.com/why-do-people-tip-generously-mostly-in-america-d654f8afcb5c?source=rss-3923c775a9ba------2</link>
            <guid isPermaLink="false">https://medium.com/p/d654f8afcb5c</guid>
            <category><![CDATA[culture]]></category>
            <category><![CDATA[psychology]]></category>
            <category><![CDATA[tipping]]></category>
            <category><![CDATA[indian]]></category>
            <category><![CDATA[tips]]></category>
            <dc:creator><![CDATA[Ameya Thorat]]></dc:creator>
            <pubDate>Fri, 22 Feb 2019 22:34:25 GMT</pubDate>
            <atom:updated>2019-02-24T15:41:06.868Z</atom:updated>
            <content:encoded><![CDATA[<h3><strong>Why don’t people tip?</strong></h3><p>During my B-school, I was fortunate to have met a lot of interesting fellow students with diverse backgrounds. I learnt more from interacting with these friends and colleagues than I did from the classroom.</p><p>During one of our Thursday evening happy hours, Jen my classmate who had waited tables in the past, asked me right up “Why are Indians such sloppy tippers?”. I was taken aback, and did not expect such a direct question that put me on the spot. I could also see a text book implementation of “selection bias” and my defense mechanisms kicked in and I immediately responded, “ how can you generalize the behavior of 1.3 billion people, based on a handful interactions ?”. Jen was a good sport and a dear friend and I knew that she was genuinely curious to find out if there is any reason I could think of that justified the sloppy tipping phenomenon, but I just was not prepared to provide an answer and we just moved on to other topics. I did however make sure that I left a generous tip at the bar that day.</p><p><strong>Gathering Data: </strong>The thought kept playing on my mind and I decided to gather more data points based on my personal interactions and observations. Over the next couple of months, every time I went out for drinks or food, I was in “Sherlock mode” . I tried hard to take a sneak peek at the tip and total amount when my friends or colleagues signed their checks. Whenever, I was not able to get a clear view, I just asked my friends about how much they tipped and made a quick note in my phone.</p><p><strong>Results:</strong> When I plotted the results and I had a WTF moment. To my surprise, I did see a pattern. I realized that Jen was partially correct in her observation, I was also relieved that it’s not just Indians, but a wider sect. While Americans on an average tipped anywhere from 12% to 22%, non-Americans maxed out at 12% and that too when they were super drunk!</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/717/1*7oiF21Bx6bP0t0ucRfMZxw.png" /></figure><p>I felt that I had defined the problem statement to my satisfaction and could now dive deeper to understand the real “why”.</p><p><strong>Explanation of the difference</strong>: Based on further conversations, debates and occasional heated exchanges, I narrowed down the behavior of sloppy tipping to three main reasons:</p><ol><li><strong>Awareness</strong>: Most people I spoke to, did not realize that the federal minimum wage for employees that receive a tip is $2.13 (except for Alaska, where it is $9.89). People assumed that the waitresses were paid the federal minimum wage in the US (which is $7.25, barring 29 states that have a minimum wage higher than the federal minimum). Once I pointed out that difference, (my) people seemed more receptive to tipping better.</li><li><strong>Culture</strong>: Imagine growing up in a country with 1.2 billion people belonging to differing dialects, religions, values, beliefs and ethnicities. A country where “diversity” can be both, a strength and a weakness, depending on the circumstances. No matter how rich or poor you are, our culture always has high regard for getting the most out of your buck. You learn to negotiate growing up, may it be the traffic cop trying to fine you for not carrying the relevant documents or the taxi driver you has rounded up the fare to the nearest higher multiple of 10, negotiations is survival. When immigrants come to the US, ingrained in their culture, they don’t care about the judgement people seem to form, all they care about is survival</li><li><strong>Exchange Rates (affordability):</strong> Today, 1 USD is ~71 Indian rupees, so all the savings are divided by a factor of 71 and the costs go up. Imagine walking into the nearest grocery store to buy bread, I am spending at least 3$ = 213 rupees for a bread that I would get for 40–50 rupees back at home, so the costs have gone up by ~400%. To top it all, you are still a student and have student loans. Tipping, is definitely least of the priorities. But as you climb the social and professional ladder, the priorities change, and you start paying attention to other social norms.</li></ol><p>To summarize, everyone needs to contribute their due share and respect the traditions/culture of the place you are at without having to compromise on the values you have been bought up with. Amalgamating the best of different cultures is the reason America has been at the epitome of innovation and we need to ensure we all keep fire burning.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=d654f8afcb5c" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>