Machine learning, Bayesian statistics, and the web

I am a Quantitative Analyst at Google, where I work on time-series analysis of web traffic, anomaly detection, Bayesian causal inference, and auction theory. Before joining Google, I was a postdoctoral researcher at the Department of Information Technology and Electrical Engineering at ETH Zurich working on Bayesian inference in high-dimensional time-series data.

What excites me about statistical data analysis is how much progress there has been in recent years in our abilities to understand large time-series datasets. Not only have there been advances in approximate inference; there are also increasingly powerful tools for running massively parallelized analyses. Some concepts, for example variational Bayesian techniques, owe their popularity in part to computationally intensive disciplines such as computational neuroscience. The same concepts have proven powerful for extracting insights from web traffic and advertising data.

Recent publications

  • New CausalImpact, a new R package for estimating causal effects in time series
    Kay H. Brodersen / Google, Inc. (2014)
    Announcement   Project site   GitHub repository   Documentation  
  • Inferring causal impact using Bayesian structural time-series models
    K.H. Brodersen, F. Gallusser, J. Koehler, N. Remy, S. Scott (2014).
    Annals of Applied Statistics
    research.google.com
  • Variational Bayesian mixed-effects inference for classification studies
    K.H. Brodersen, J. Daunizeau, C. Mathys, J.R. Chumbley, J.M. Buhmann, K.E. Stephan (2013).
    NeuroImage
    PDF

Show all publications