Projected semi-stochastic gradient descent method with mini-batch scheme under weak strong convexity assumption
From MaRDI portal
Publication:1695084
DOI10.1007/978-3-319-66616-7_7zbMath1383.93096arXiv1612.05356MaRDI QIDQ1695084
Publication date: 6 February 2018
Full work available at URL: https://arxiv.org/abs/1612.05356
linear convergence; variance reduction; stochastic gradient; support vector machine (SVM); weak strong convexity
90C15: Stochastic programming
93E20: Optimal stochastic control
49N15: Duality theory (optimization)
49J55: Existence of optimal solutions to problems involving randomness