Alex RobertsinTowards Data ScienceExploring Best Test Size, Number of Folds and Repeated Hold-OutStudy gains in guaranteed, supported and mean accuracy; discover biasesSep 9, 20221Sep 9, 20221
Alex RobertsA Frequentist modification of the logistic function for probability estimates of small datasetsWhy exponential tails can be dangerous and how to limit their range.May 30, 2022May 30, 2022
Alex RobertsA different way to think about logistic regressionUsually, the motivation of logistic regression as a likelihood function is not really talked about. We are just told that we have to do…May 26, 2022May 26, 2022
Alex RobertsThe secretary problem — math, insights and statisticsThe secretary problem is over 70 years old. It formulates the rules for a game as follows: you are to hire a secretary and have n…May 21, 2022May 21, 2022
Alex RobertsHow to teach current in a wire (or how not to)If you’re going to teach how a basic circuit works, please never again give the analogy of water flowing through a pipe. Making that…Sep 1, 2018Sep 1, 2018