Deep Relaxation: partial differential equations for optimizing deep neural networks

April 17, 2017 ยท Declared Dead ยท ๐Ÿ› Research in the Mathematical Sciences

๐Ÿ‘ป CAUSE OF DEATH: Ghosted
No code link whatsoever

"No code URL or promise found in abstract"

Evidence collected by the PWNC Scanner

Authors Pratik Chaudhari, Adam Oberman, Stanley Osher, Stefano Soatto, Guillaume Carlier arXiv ID 1704.04932 Category cs.LG: Machine Learning Cross-listed math.AP, math.OC Citations 159 Venue Research in the Mathematical Sciences Last Checked 4 months ago
Abstract
In this paper we establish a connection between non-convex optimization methods for training deep neural networks and nonlinear partial differential equations (PDEs). Relaxation techniques arising in statistical physics which have already been used successfully in this context are reinterpreted as solutions of a viscous Hamilton-Jacobi PDE. Using a stochastic control interpretation allows we prove that the modified algorithm performs better in expectation that stochastic gradient descent. Well-known PDE regularity results allow us to analyze the geometry of the relaxed energy landscape, confirming empirical evidence. The PDE is derived from a stochastic homogenization problem, which arises in the implementation of the algorithm. The algorithms scale well in practice and can effectively tackle the high dimensionality of modern neural networks.
Community shame:
Not yet rated
Community Contributions

Found the code? Know the venue? Think something is wrong? Let us know!

๐Ÿ“œ Similar Papers

In the same crypt โ€” Machine Learning

Died the same way โ€” ๐Ÿ‘ป Ghosted