Nonexpansive Operators for Convex Optimization Algorithms

Nonexpansive operators can explain many convex optimization algorithms. In this talk, I will introduce some nonexpansive operators and use them to show the convergence of many convex optimization algorithms, including gradient descent, proximal gradient descent, Douglas-Rachford splitting, Davis-Yin three-operator splitting, primal-dual algorithm, etc. This talk is accessible to general audience even without a convex optimization background.  

学术海报.pdf