Implicit dynamic regularization in deep networks
Author(s)
Poggio, Tomaso; Liao, Qianli
DownloadCBMM-Memo-112.pdf (2.289Mb)
Additional downloads
Metadata
Show full item recordAbstract
Square loss has been observed to perform well in classification tasks, at least as well as crossentropy. However, a theoretical justification is lacking. Here we develop a theoretical analysis for the square loss that also complements the existing asymptotic analysis for the exponential loss.
Date issued
2020-08-17Publisher
Center for Brains, Minds and Machines (CBMM)
Series/Report no.
CBMM Memo;112