MLLGJul 4, 2017

Learning Deep Energy Models: Contrastive Divergence vs. Amortized MLE

arXiv:1707.00797v129 citations
Originality Synthesis-oriented
AI Analysis

This work addresses improving model learning in machine learning, but it appears incremental as it builds on existing methods without a clear breakthrough.

The paper tackled the problem of learning deep energy models by proposing new algorithms, showing that SteinCD achieves good test likelihood and SteinGAN generates realistic images, suggesting a combination of GAN-style methods with energy-based learning.

We propose a number of new algorithms for learning deep energy models and demonstrate their properties. We show that our SteinCD performs well in term of test likelihood, while SteinGAN performs well in terms of generating realistic looking images. Our results suggest promising directions for learning better models by combining GAN-style methods with traditional energy-based learning.

Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes