Statistical Learning
DistributionallyRobustOptimizationwithData Geometry
It has been well documented that high dimensional data approximately resides on low dimensional manifolds.
Statistical Optimality of Stochastic Gradient Descent on Hard Learning Problems through Multiple Passes
Loucas Pillaud-Vivien, Alessandro Rudi, Francis Bach
Neural Information Processing Systems http://nips.cc/
Hessian-based Analysis of Large Batch Training and Robustness to Adversaries
Zhewei Yao, Amir Gholami, Qi Lei, Kurt Keutzer, Michael W. Mahoney
Neural Information Processing Systems http://nips.cc/
min
LetAbean nHermitian matrixandletBbea(n 1) (n 1)matrixwhich is constructed by deleting thei-th row andi-th column ofA. Denote thatฮฆ = [ฯ(x1),...,ฯ(xn)] Rn D, where D is the dimension of feature spaceH. Performing rank-n singular value decomposition (SVD) onฮฆ, we have ฮฆ = HฮฃV, where H Rn n, ฮฃ Rn n is a diagonal matrix whose diagonal elements are the singular values of ฮฆ,andV RD n. F(ฮฑ) in Eq.(21) is proven differentiable and thep-th component of the gradient is F(ฮฑ) ฮฑp = Then, a reduced gradient descent algorithm [26] is adopted to optimize Eq.(21). The three deep neural networks are pre-trained on the ImageNet[5].
Nonparametric Bayesian Lomax delegate racing for survival analysis with competing risks
Neural Information Processing Systems http://nips.cc/
Random Feature Stein Discrepancies
Jonathan Huggins, Lester Mackey
Neural Information Processing Systems http://nips.cc/