Borun Chowdhury Ph.D.Gambler’s Problem — When Inaction Is Infact OptimalThe Gambler’s Problem from Barto and Sutton has attracted a lot of discussion and confusion. I give a seemingly paradoxical and ultimately…Dec 21, 2022Dec 21, 2022
Borun Chowdhury Ph.D.Monty Hall Done RightThe Monty Hall problem is solved with an arbitrary number of cars and goats using conditional probabilities. As a bonus, I include a colab…Dec 8, 2021Dec 8, 2021
Borun Chowdhury Ph.D.inTowards Data ScienceBilliards with Fred the Frequentist and Bayer the Bayesian [Bayer wins :)]I discuss here an experiment due to Bayes himself that highlights the beauty, elegance and CORRECTNESS of Bayesian approach even as the…Oct 9, 20214Oct 9, 20214
Borun Chowdhury Ph.D.Do you understand Newton’s Laws?Many do not know Newton’s Laws of motion but think they doOct 5, 20211Oct 5, 20211
Borun Chowdhury Ph.D.inTowards Data ScienceUnderstanding classification metricsFor a simple binary classification task the beginner is often confused by a ton of metrics. Here I give a bottom up explanation of…Jun 10, 2021Jun 10, 2021
Borun Chowdhury Ph.D.inTowards Data ScienceClass weights for categorical lossMay 22, 2021May 22, 2021
Borun Chowdhury Ph.D.inTowards Data ScienceThe Cauchy Distribution — Your distribution so fat…The Cauchy distribution is so fat tailed that it does not have a mean or variance. We look at a physically motivated example and its…Dec 28, 2020Dec 28, 2020
Borun Chowdhury Ph.D.inThe StartupWhy Most Articles on the Central Limit Theorem Are MisleadingMost articles say that sample means of random variables drawn from any distribution tends to Gaussians but this is incorrectDec 26, 20201Dec 26, 20201
Borun Chowdhury Ph.D.inTowards Data ScienceMaking Docker and Conda play well togetherBuild a Docker image running a multi-kernel Jupyter notebook server in fifteen minutesDec 26, 2020Dec 26, 2020
Borun Chowdhury Ph.D.inTowards Data ScienceCustom metrics in Keras and how simple they are to use in tensorflow2.2Keras has simplified DNN based machine learning a lot and it keeps getting better. Here we show how to implement metric based on the…Jul 2, 20201Jul 2, 20201