AISTATS 2019poster2 citations

A recurrent Markov state-space generative model for sequences

Anand Ramachandran, Steve Lumetta, Eric Klee, Deming Chen

Abstract

While the Hidden Markov Model (HMM) is a versatile generative model of sequences capable of performing many exact inferences efficiently, it is not suited for capturing complex long-term structure in the data. Advanced state-space models based on Deep Neural Networks (DNN) overcome this limitation but cannot perform exact inferences. In this article, we present a new generative model for sequences that combines both aspects, the ability to perform exact inferences and the ability to model long-term structure, by augmenting the HMM with a deterministic, continuous state variable modeled through a Recurrent Neural Network. We empirically study the performance of the model on (i) synthetic data comparing it to the HMM, (ii) a supervised learning task in bioinformatics where it outperforms two DNN-based regressors and (iii) in the generative modeling of music where it outperforms many prominent DNN-based generative models.

BibTeX
@InProceedings{pmlr-v89-ramachandran19a,
  title = 	 {A recurrent Markov state-space generative model for sequences},
  author =       {Ramachandran, Anand and Lumetta, Steve and Klee, Eric and Chen, Deming},
  booktitle = 	 {Proceedings of the Twenty-Second International Conference on Artificial Intelligence and Statistics},
  pages = 	 {3070--3079},
  year = 	 {2019},
  editor = 	 {Chaudhuri, Kamalika and Sugiyama, Masashi},
  volume = 	 {89},
  series = 	 {Proceedings of Machine Learning Research},
  month = 	 {16--18 Apr},
  publisher =    {PMLR},
  pdf = 	 {http://proceedings.mlr.press/v89/ramachandran19a/ramachandran19a.pdf},
  url = 	 {https://proceedings.mlr.press/v89/ramachandran19a.html},
  abstract = 	 {While the Hidden Markov Model (HMM) is a versatile generative model of sequences capable of performing many exact inferences efficiently, it is not suited for capturing complex long-term structure in the data.  Advanced state-space models based on Deep Neural Networks (DNN) overcome this limitation but cannot perform exact inferences.  In this article, we present a new generative model for sequences that combines both aspects, the ability to perform exact inferences and the ability to model long-term structure, by augmenting the HMM with a deterministic, continuous state variable modeled through a Recurrent Neural Network. We empirically study the performance of the model on (i) synthetic data comparing it to the HMM, (ii) a supervised learning task in bioinformatics where it outperforms two DNN-based regressors and (iii) in the generative modeling of music where it outperforms many prominent DNN-based generative models.}
}
A recurrent Markov state-space generative model for sequences · AISTATS 2019