Exploiting negative curvature in deterministic and stochastic optimization
Exploiting negative curvature in deterministic and stochastic optimization is a scholarly work, published in 2018 in ''Mathematical Programming''. The main subjects of the publication include convergence, stochastic optimization, mathematics, computer science, descent, Bounding overwatch, curvature, mathematical optimization, biological function, compressed sensing, machine learning, stochastic gradient descent, and algorithm. The authors present new frameworks for combining descent and negative curvature directions: alternating two-step approaches and dynamic step approaches.