Deep Conditional Measure Quantization

Abstract

Quantization of a probability measure means representing it with a finite set of Dirac masses that approximates the input distribution well enough (in some metric space of probability measures). Various methods exists to do so, but the situation of quantizing a conditional law has been less explored. We propose a method, called DCMQ, involving a Huber-energy kernel-based approach coupled with a deep neural network architecture. The method is tested on several examples and obtains promising results.

References in corpus (5)

• Neural Discrete Representation Learning
• Efficient Training of Language Models to Fill in the Middle
• Optimal quantization of the mean measure and applications to statistical learning
• Vector Quantized Wasserstein Auto-Encoder
• A Deep Generative Approach to Conditional Sampling