the archive
#first-order methods
2 resultsmath.OC2019
Some Worst-Case Datasets of Deterministic First-Order Methods for Solving Binary Logistic Regression
Yuyuan Ouyang, Trevor Squires
The paper constructs specific worst-case datasets that force deterministic first-order algorithms for binary logistic regression to need at least O(1/√ε) oracle calls to achieve an…
#binary logistic regression#first-order methods#worst-case analysis#oracle complexity
math.OC2019
Improved Oracle Complexity of Variance Reduced Methods for Nonsmooth Convex Stochastic Composition Optimization
Tianyi Lin, Chenyou Fan, Mengdi Wang
The paper analyzes stochastic compositional variance reduced gradient methods for nonsmooth convex composition problems and proves improved incremental first-order oracle complexit…
#stochastic optimization#variance reduction#nonsmooth convex#compositional optimization