%0 Journal Article %1 kawaguchi2017generalization %A Kawaguchi, Kenji %A Kaelbling, Leslie Pack %A Bengio, Yoshua %D 2017 %K bounds deep-learning generalization learning machine-learning mathematics probability stats theory %T Generalization in Deep Learning %U http://arxiv.org/abs/1710.05468 %X This paper provides non-vacuous and numerically-tight generalization guarantees for deep learning, as well as theoretical insights into why and how deep learning can generalize well, despite its large capacity, complexity, possible algorithmic instability, nonrobustness, and sharp minima, responding to an open question in the literature. We also propose new open problems and discuss the limitations of our results.

@article{kawaguchi2017generalization, abstract = {This paper provides non-vacuous and numerically-tight generalization guarantees for deep learning, as well as theoretical insights into why and how deep learning can generalize well, despite its large capacity, complexity, possible algorithmic instability, nonrobustness, and sharp minima, responding to an open question in the literature. We also propose new open problems and discuss the limitations of our results.}, added-at = {2019-05-31T16:41:12.000+0200}, author = {Kawaguchi, Kenji and Kaelbling, Leslie Pack and Bengio, Yoshua}, biburl = {https://www.bibsonomy.org/bibtex/2040f7d26bd2026fa39dc168d9ef8f987/kirk86}, description = {[1710.05468] Generalization in Deep Learning}, interhash = {73e5b6798e695efd53255d7740ddaf8f}, intrahash = {040f7d26bd2026fa39dc168d9ef8f987}, keywords = {bounds deep-learning generalization learning machine-learning mathematics probability stats theory}, note = {cite arxiv:1710.05468Comment: To appear in Mathematics of Deep Learning, Cambridge University Press. All previous results remain unchanged}, timestamp = {2019-05-31T16:41:12.000+0200}, title = {Generalization in Deep Learning}, url = {http://arxiv.org/abs/1710.05468}, year = 2017 }