4 code implementations • 2 Mar 2016 • Alp Kucukelbir, Dustin Tran, Rajesh Ranganath, Andrew Gelman, David M. Blei
Probabilistic modeling is iterative.
6 code implementations • 4 Jan 2016 • David M. Blei, Alp Kucukelbir, Jon D. McAuliffe
One of the core problems of modern statistics is to approximate difficult-to-compute probability densities.
1 code implementation • 2 Nov 2014 • Alp Kucukelbir, David M. Blei
We develop population empirical Bayes (POP-EB), a hierarchical framework that explicitly models the empirical population distribution as part of Bayesian analysis.
1 code implementation • ICML 2017 • Yixin Wang, Alp Kucukelbir, David M. Blei
We propose a way to systematically detect and mitigate mismatch of a large class of probabilistic models.
1 code implementation • 30 Nov 2017 • Yunhao Tang, Alp Kucukelbir
We propose a framework that directly tackles the probability distribution of the value function parameters in Deep Q Network (DQN), with powerful variational inference subroutines to approximate the posterior of the parameters.
no code implementations • 31 Oct 2016 • Dustin Tran, Alp Kucukelbir, Adji B. Dieng, Maja Rudolph, Dawen Liang, David M. Blei
Probabilistic modeling is a powerful approach for analyzing empirical information.
no code implementations • 24 May 2016 • Alp Kucukelbir, David M. Blei
We propose to evaluate a model through posterior dispersion.
no code implementations • NeurIPS 2015 • Alp Kucukelbir, Rajesh Ranganath, Andrew Gelman, David M. Blei
With ADVI we can use variational inference on any model we write in Stan.
no code implementations • ICML 2017 • Alp Kucukelbir, Yixin Wang, David M. Blei
We propose to evaluate a model through posterior dispersion.
no code implementations • 29 May 2019 • Yunhao Tang, Krzysztof Choromanski, Alp Kucukelbir
Evolution Strategies (ES) are a powerful class of blackbox optimization techniques that recently became a competitive alternative to state-of-the-art policy gradient (PG) algorithms for reinforcement learning (RL).
no code implementations • 13 Jun 2020 • Yunhao Tang, Alp Kucukelbir
We propose a graphical model framework for goal-conditioned RL, with an EM algorithm that operates on the lower bound of the RL objective.