Predictive State Smoothing (PRESS): Scalable non-parametric regression

Georg M. Goerg (https://research.google.com/pubs/GeorgGoerg.html)from Google will be talking about his work on predictive state smoothing (PRESS)

Abstract

We introduce predictive state smoothing (PRESS), a novel semi-parametric regression technique for high-dimensional data using predictive state representations. PRESS is a fully probabilistic model for the optimal kernel smoothing matrix. We present efficient algorithms for the joint estimation of the state space as well as the non-linear mapping of observations to predictive states and as an alternative algorithms to minimize leave-one-out cross-validation error. The proposed estimator is straightforward to implement using (stochastic) gradient descent and scales well for large N and large p. LASSO penalty parameters as well the optimal smoothness can be estimated as part of the optimization. Finally, we show that out-of-sample predictions are on par with or better than alternative state-of-the-art regression methods on the abalone and MNIST benchmark datasets. Yet unlike alternative methods PRESS gives meaningful domain-specific insights and can be used for statistical inference via regression coefficients.

Directions

Chelsea Market Entrance at 75 Ninth Ave, New York, NY 10011 (BTWN 15th & 16th street) Once you enter the building there are 2 elevators on the immediate right hand side (near Security station) take either elevator up to the 4th floor. Elevators require a badge, so someone from Google will have to escort you & guests to 4th floor. We have Wildwood conf. room for the event.