Saddlepoints in Unsupervised Least Squares
This paper sheds light on the risk landscape of unsupervised least squares in the context of deep auto-encoding neural nets. We formally establish an equivalence between unsupervised least squares and principal manifolds. This link provides insight into the risk landscape of auto--encoding under the mean squared error, in particular all non-trivial critical points are saddlepoints. Finding saddlepoints is in itself difficult, overcomplete auto-encoding poses the additional challenge that the saddlepoints are degenerate. Within this context we discuss regularization of auto-encoders, in particular bottleneck, denoising and contraction auto-encoding and propose a new optimization strategy that can be framed as particular form of contractive regularization.
Apr-11-2021
- Country:
- Europe > United Kingdom
- England > Cambridgeshire > Cambridge (0.04)
- Africa > Middle East
- Tunisia > Ben Arous Governorate > Ben Arous (0.04)
- Europe > United Kingdom
- Genre:
- Research Report (0.64)
- Technology: