Off-policy evaluation in infinite-horizon reinforcement learning with latent confounders

A Bennett, N Kallus, L Li… - … Conference on Artificial …, 2021 - proceedings.mlr.press
Off-policy evaluation (OPE) in reinforcement learning is an important problem in settings
where experimentation is limited, such as healthcare. But, in these very same settings …

Predictive state recurrent neural networks

C Downey, A Hefny, B Boots… - Advances in Neural …, 2017 - proceedings.neurips.cc
We present a new model, Predictive State Recurrent Neural Networks (PSRNNs), for filtering
and prediction in dynamical systems. PSRNNs draw on insights from both Recurrent Neural …

Anchored discrete factor analysis

Y Halpern, S Horng, D Sontag - arxiv preprint arxiv:1511.03299, 2015 - arxiv.org
We present a semi-supervised learning algorithm for learning discrete factor analysis
models with arbitrary structure on the latent variables. Our algorithm assumes that every …

Improving predictive state representations via gradient descent

N Jiang, A Kulesza, S Singh - Proceedings of the AAAI Conference on …, 2016 - ojs.aaai.org
Predictive state representations (PSRs) model dynamical systems using appropriately
chosen predictions about future observations as a representation of the current state. In …

Practical learning of predictive state representations

C Downey, A Hefny, G Gordon - arxiv preprint arxiv:1702.04121, 2017 - arxiv.org
Over the past decade there has been considerable interest in spectral algorithms for
learning Predictive State Representations (PSRs). Spectral algorithms have appealing …

Fast and consistent learning of hidden Markov models by incorporating non-consecutive correlations

R Mattila, C Rojas, E Moulines… - International …, 2020 - proceedings.mlr.press
Can the parameters of a hidden Markov model (HMM) be estimated from a single sweep
through the observations–and additionally, without being trapped at a local optimum in the …

Latent feature lasso

IEH Yen, WC Lee, SE Chang… - International …, 2017 - proceedings.mlr.press
The latent feature model (LFM), proposed in (Griffiths\& Ghahramani, 2005), but possibly
with earlier origins, is a generalization of a mixture model, where each instance is generated …

Feature-weighted survival learning machine for COPD failure prediction

J Zhang, S Wang, J Courteau, L Chen, G Guo… - Artificial Intelligence in …, 2019 - Elsevier
Chronic obstructive pulmonary disease (COPD) yields a high rate of failures such as
hospital readmission and death in the United States, Canada and worldwide. COPD failure …

Completing state representations using spectral learning

N Jiang, A Kulesza, S Singh - Advances in Neural …, 2018 - proceedings.neurips.cc
A central problem in dynamical system modeling is state discovery—that is, finding a
compact summary of the past that captures the information needed to predict the future …

Identification of hidden Markov models using spectral learning with likelihood maximization

R Mattila, CR Rojas, V Krishnamurthy… - 2017 IEEE 56th …, 2017 - ieeexplore.ieee.org
In this paper, we consider identifying a hidden Markov model (HMM) with the purpose of
computing estimates of joint and conditional (posterior) probabilities over observation …