Pith. sign in

Self bounding learning algorithms

1 Pith paper cite this work. Polarity classification is still indexing.

1 Pith paper citing it

fields

cs.LG 1

years

2019 1

verdicts

REJECT 1

representative citing papers

PAC-Bayes with Backprop

cs.LG · 2019-08-19 · reject · novelty 5.0

Training neural networks with PAC-Bayes objectives yields MNIST test error of 1.4% and a non-vacuous risk bound of 2.3%, much tighter than prior PAC-Bayes certificates.

citing papers explorer

Showing 1 of 1 citing paper.

  • PAC-Bayes with Backprop cs.LG · 2019-08-19 · reject · none · ref 8

    Training neural networks with PAC-Bayes objectives yields MNIST test error of 1.4% and a non-vacuous risk bound of 2.3%, much tighter than prior PAC-Bayes certificates.