Understanding the Limitations of Variational Mutual Information Estimators

By Jiaming Song, Stefano Ermon. 2019

248 citations recorded by the source corpus.

Variational approaches based on neural networks are showing promise for estimating mutual information (MI) between high dimensional variables. However, they can be difficult to use in practice due to poorly understood bias/variance tradeoffs. We theoretically show that, under some conditions, estimators such as MINE exhibit variance that could grow exponentially with the true amount of underlying MI. We also empirically demonstrate that existing estimators fail to satisfy basic self-consistency properties of MI, such as data processing and additivity under independence. Based on a unified perspective of variational approaches, we develop a new estimator that focuses on variance reduction. Empirical results on standard benchmark tasks demonstrate that our proposed estimator exhibits improved bias-variance trade-offs on standard benchmark tasks.

Subfields: Artificial Intelligence, Probabilistic & Bayesian Methods, Information Theory

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

  1. Mohit Bansal
  2. Ruslan Salakhutdinov
  3. Percy Liang
  4. Jianfeng Gao
  5. Yoshua Bengio
  6. Sergey Levine
  7. Stefano Ermon
  8. Zhe Gan

In Probabilistic & Bayesian Methods

  1. Stefano Ermon
  2. Jiaming Song
  3. Arash Vahdat
  4. Shengjia Zhao
  5. Jan Kautz
  6. Lawrence Carin
  7. Zhe Gan
  8. Honglak Lee

In Information Theory

  1. Stefano Ermon
  2. Jiaming Song
  3. Zhe Gan
  4. Lawrence Carin
  5. Ruslan Salakhutdinov
  6. Yao-Hung Hubert Tsai
  7. Honglak Lee
  8. Polina Golland

Citations in the corpus