Machine Learning Data Analysis Project Presentation / Heinz School First Paper

We present a scalable Gaussian process model for identifying and characterizing smooth multidimensional changepoints, and automatically learning changes in expressive covariance structure. We use Random Kitchen Sink features to flexibly define a change surface in combination with expressive spectral mixture kernels to capture the complex statistical structure. Through the use of novel methods for additive non-separable kernels, we scale the model to large datasets. We demonstrate the model on numerical simulations as well as applying it to real world spatio-temporal data. Specifically, we model state level incidence rates of measles in the United States both before and after the introduction of the measles vaccine. Additionally we model zip code level requests for lead testing kits in New York City over the past two years in the midst of heightened concerns…


Link to Full Article: Machine Learning Data Analysis Project Presentation / Heinz School First Paper