Stochastic variational inference for hidden Markov models

Authors: Nicholas Foti, Jason Xu, Dillon Laird, Emily B. Fox

NeurIPS 2014 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We demonstrate the effectiveness of our algorithm on synthetic experiments and a large genomics dataset where a batch algorithm is computationally infeasible.
Researcher Affiliation Academia Nicholas J. Foti , Jason Xu , Dillon Laird, and Emily B. Fox University of Washington {nfoti@stat,jasonxu@stat,dillonl2@cs,ebfox@stat}.washington.edu
Pseudocode Yes Algorithm 1 Stochastic Variational Inference for HMMs (SVIHMM) and Algorithm 2 Grow Buf procedure.
Open Source Code No The paper does not explicitly state that source code is provided or offer a link to a repository.
Open Datasets Yes We apply the SVIHMM algorithm to a massive human chromatin dataset provided by the ENCODE project [24]. [24] ENCODE Project Consortium. An integrated encyclopedia of DNA elements in the human genome. Nature, 489(7414):57 74, September 2012.
Dataset Splits Yes In Fig. 1(b), we see similar trends in terms of predictive log-probability holding out 10% of the observations as a test set and using 5-fold cross validation.
Hardware Specification No The paper does not specify the exact hardware (e.g., CPU/GPU models, memory) used for running the experiments.
Software Dependencies No The paper does not specify software dependencies with version numbers.
Experiment Setup Yes For each per parameter setting, we ran 20 random restarts of SVIHMM for 100 iterations and batch VB until convergence of the ELBO. A forgetting rate κ parametrizes step sizes ρn = (1 + n) κ. We fix the total number of observations L M used per iteration of SVIHMM such that increasing M implies decreasing L (and vice versa).