On the Information Theoretic Distance Measures and Bidirectional Helmholtz Machines
arXiv:1807.06054v1
Originality Incremental advance
AI Analysis
This work addresses a specific issue in machine learning for researchers, but it appears incremental as it builds on existing Helmholtz machines with theoretical extensions.
The authors tackled the problem of improving shallow bidirectional Helmholtz machines by connecting them to information theory, resulting in a generalized model that substantially outperforms previous versions.
By establishing a connection between bi-directional Helmholtz machines and information theory, we propose a generalized Helmholtz machine. Theoretical and experimental results show that given \textit{shallow} architectures, the generalized model outperforms the previous ones substantially.