Abstract
Many machine learning tasks, such as learning with invariance and policy evaluation in reinforcement learning, can be characterized as problems of learning from conditional distributions. In such problems, each sample x itself is associated with a conditional distribution p(z|x) represented by samples {zi}Mi=1, and the goal is to learn a function f that links these conditional distributions to target values y. These problems become very challenging when we only have limited samples or in the extreme case only one sample from each conditional distribution. Commonly used approaches either assume that z is independent of x, or require an overwhelmingly large set of samples from each conditional distribution. To address these challenges, we propose a novel approach which employs a new min-max reformulation of the learning from conditional distribution problem. With such new reformulation, we only need to deal with the joint distribution p(z, x). We also design an efficient learning algorithm, Embedding-SGD, and establish theoretical sample complexity for such problems. Finally, our numerical experiments, on both synthetic and real-world datasets, show that the proposed approach can significantly improve over existing algorithms.
Original language | English (US) |
---|---|
State | Published - 2017 |
Event | 20th International Conference on Artificial Intelligence and Statistics, AISTATS 2017 - Fort Lauderdale, United States Duration: Apr 20 2017 → Apr 22 2017 |
Conference
Conference | 20th International Conference on Artificial Intelligence and Statistics, AISTATS 2017 |
---|---|
Country/Territory | United States |
City | Fort Lauderdale |
Period | 4/20/17 → 4/22/17 |
ASJC Scopus subject areas
- Artificial Intelligence
- Statistics and Probability