To help students get a better feel for three of the most popular “multi-armed bandit” exploration/exploitation balancing strategies (Epsilon Greedy, Thompson Sampling, and Upper Confidence Bound), I combined my R package “contextual” with the versatile …
Algorithmic Art: The Bubble
The Bubble, an algorithmic artwork I have been working on with the artist and photographer Johan Nieuwenhuize, was unveiled today during the opening of the academic year at the The Hague University of Applied Sciences. …
Music as Medicine: Sync
Above, a video visualizing two of our participants’ ECG responding to music in real-time. Yellow and green lines represent their ECG’s, red the level of synchrony between them. The audio is a live recording …
Shrinkage in Contextual Bandits with Dependent Observations
Jules Kruijswijk (first author), Maurits Kaptein and Florian Böing-Messing and I have recently been working on a paper on the application of partial pooling (or shrinkage) models in (contextual) multi-armed bandit problems: The (contextual) multi-armed …
StreamingBandit UI
Recently, I have completed a Javascript (ES6, React, and Material Design) admin application for “StreamingBandit”, a Python server that helps setting up and running online streaming (contextual) bandit experiments. This management front-end makes it very easy …