Dropout is good for more than regularization // outlier detection, adversarial rejection

Dropout is good for more than regularization // outlier detection, adversarial rejection

Time point change detection the bayesian way

Quantifying uncertainty in a regression model

Quantifying uncertainty in machine learning models using ensemble methods

Unpacking the winners and losers from the recession

Taking a look at the history of rome with D3.js and Python

Making smooth, animated dataviz projects with python

Which is the Worst Airport in America?

How has LEGO quantitatively changed over time?

What does the natural world look like on average?

The bins and scales we use determine the stories we tell

Lots of actors were hedging their bets this year

Did segregation really end a generation ago? Here we investigate how it continues, quantitatively.

Discount bioinformatics with premium math and dumpster dived data

What tools from signals and systems analysis can help us understand trends in the political process

Ever wonder if giving congress people term limits would be a good idea? Here, I use logistic regression to look at a few landmark pieces of legislation and determine what role (if any) limiting term limits would play on the outcome of the vote

Are your country’s smokers casual or hardcore? Here, I investigate the differences in prevalence of smoking and how many cigarettes those smokers consume among different nations.

Pizzas, how do they work? Well think long and hard before you order a pizza in the wrong size or the wrong state!