Learning to be Global Optimizer

Abstract

The advancement of artificial intelligence has cast a new light on the development of optimization algorithm. This paper proposes to learn a two-phase (including a minimization phase and an escaping phase) global optimization algorithm for smooth non-convex functions. For the minimization phase, a model-driven deep learning method is developed to learn the update rule of descent direction, which is formalized as a nonlinear combination of historical information, for convex functions. We prove that the resultant algorithm with the proposed adaptive direction guarantees convergence for convex functions. Empirical study shows that the learned algorithm significantly outperforms some well-known classical optimization algorithms, such as gradient descent, conjugate descent and BFGS, and performs well on ill-posed functions. The escaping phase from local optimum is modeled as a Markov decision process with a fixed escaping policy. We further propose to learn an optimal escaping policy by reinforcement learning. The effectiveness of the escaping policies is verified by optimizing synthesized functions and training a deep neural network for CIFAR image classification. The learned two-phase global optimization algorithm demonstrates a promising global search capability on some benchmark functions and machine learning tasks.

Distinctive properties of biological neural networks and recent advances in bottom-up approaches toward a better biologically plausible neural network

“…An orchestrated strategy is required for these local processes of plasticity to result in the learning of a certain function. Learning is the adaptation of a neural network to approximate a function that maps from the input from the environment to the target output, which is a global optimization process (Zhang H. et al, 2020). The optimization target and the algorithm for efficiently reaching the target by combining local processes should be elucidated to define this optimization.…”