Statistical Learning
Competitive Gradient Descent
Florian Schaefer, Anima Anandkumar
Neural Information Processing Systems http://nips.cc/
Implicit Bias of Gradient Descent on Linear Convolutional Networks
Suriya Gunasekar, Jason D. Lee, Daniel Soudry, Nati Srebro
Consequently, optimization objectives for learning such high capacity models have many global minima that fit training data perfectly.
PRIOR: Personalized Prior for Reactivating the Information Overlooked in Federated Learning
Bregman divergence for greater adaptability in personalized scenarios.
Demystifying Black-box Models with Symbolic Metamodels
Ahmed M. Alaa, Mihaela van der Schaar
Typically,G would be set as the class of all arithmetic, polynomial, algebraic, closed-form, or analytic expressions.
Total stochastic gradient algorithms and applications in reinforcement learning
Neural Information Processing Systems http://nips.cc/