While I can't talk about the mechanics of most of my work, I love to share the general concepts of what I've learned on my most recent projects. My blog posts are mostly educational in nature and discuss practical and theoretical topics that I find interesting.
While collaborating of a pair of reinforcment learning models for digital marketing work, a collegue and I found ourselves explaining our work in experimentation to many of our business collegues. THis blog is for business analysts interested in experimentation.Read More
I've spent a lot of the last two weeks working on optimizers for a suite of reinforcement learning models I built and deployed over the last year. While PSO did not end up in any of the production models I deployed, it's one of my favorite methods to think about and discuss.Read More
Over the past few months I've been actively exploring the wide range of neural network flavors. By this I mean feed foward MLPs, CNNs, RNNs (GRU and LSTM), GANs, Seq2Seq, etc. I've collected a lessons learned about how to get these networks working when you hit a road block...and in my experimentation I hit a variety of road blocks!Read More
As many of my collegues know, I'm a huge fan of the cult of bayes. One of the reasons for my fascination with bayesian stats is it's ability to intuitively quanitfy measurement uncertainties. Here I explore methods to quanitfy model uncertainty when making predictions.Read More