Category: Data Science

John Mount reviews a book:

First the conclusion: this is a well researched and important book. My rating is a strong buy, and Bernoulli’s Fallacy is already influencing how I approach my work.

My initial “judge the book by its back cover” impression of Bernoulli’s Fallacy was negative. The back cover writes some very large checks that I was initially (and wrongly) doubtful that “its fists could cash.” The thesis is that frequentist statistics (the dominant statistical practice) is far worse than is publicly admitted, and that Bayesian methods are the fix. However, other reviews and the snippets by people I respect (such as Andrew Gelman and Persi Diaconis) convinced me to buy and read the book. And I am glad that I read it. The back cover was, in my revised opinion, fully justified.

Read on for John’s full review of a book that is quite critical of frequentist statistics in favor of Bayesian statistics—so that already makes the book a winner for me.

Comments closed

Hoen Nguyen explains a couple of terms:

In the world of search engines and data retrieval, achieving high accuracy and relevance in the results is a constant challenge. One of the techniques used to improve search results is Fuzzy Search.

This blog post will delve into the concept of fuzzy search, its implementation using the Levenshtein Distance, and how to test its effectiveness.

Levenshtein distance is also one of the techniques spell checkers use, comparing word not in its dictionary to other words within a certain distance.

Comments closed

John Cook talks philosophy of statistics:

Suppose you’re running an A/B test to determine whether a web page produces more sales with one graphic versus another. You plan to randomly assign image A or B to 1,000 visitors to the page, but after only randomizing 500 visitors you want to look at the data. Is this OK or not?

John also has a follow-up article:

Suppose you design an experiment, an A/B test of two page designs, randomizing visitors to Design A or Design B. You planned to run the test for 800 visitors and you calculated some confidence level α for your experiment.

You decide to take a peek at the data after only 300 randomizations, even though your statistician warned you in no uncertain terms not to do that. Something about alpha spending.

You can’t unsee what you’ve seen. Now what?

Read on for a very interesting discussion of the topic. I’m definitely in the Bayesian camp: learn quickly update frequently, particularly early on when you have little information on the topic and the marginal value of learning one additional piece of information is so high.

Comments closed

Matthew Mayo shares a few tips:

Imperfect data is the norm rather than the exception in machine learning. Comparably common is the binary class imbalance when the classes in a trained data remains majority/minority class, or is moderately skewed. Imbalanced data can undermine a machine learning model by producing model selection biases. Therefore in the interest of model performance and equitable representation, solving the problem of imbalanced data during training and evaluation is paramount.

This article will define imbalanced data, resampling strategies as solution, appropriate evaluation metrics, kinds of algorithmic approaches, and the utility of synthetic data and data augmentation to address this imbalance.

Read on for five recommendations, starting with what you should know and then offering up four options for what you can do.

Comments closed