Bridging Black-Box Optimization and Machine Learning for Dynamic Algorithm Configuration 

The choice of a best-possible solver for a given optimization problem crucially depends on the problem characteristics and on the available computational resources. Automated Machine Learning (AutoML) aids in selecting and configuring this solver. Until very recently, however, AutoML approaches were restricted to a static selection of algorithm and configuration, thereby ignoring that the best choices can drastically change during the optimization process, e.g., by requiring exploration in the early phases while preferring exploitation as more information about the problem at hand becomes available. Dynamic Algorithm Configuration (DAC) addresses this shortcoming by explicitly asking for a state-dependent selection of algorithms and configurations. The Opt4DAC project will bring together experts in Optimization and in AutoML to develop approaches that identify suitable switches between different solvers and their configurations on the fly. To this end, we aim for (i) warm-starting optimizers with previous optimization runs, (ii) designing relevant features to describe the current state of the optimization process, and (iii) mechanisms to identify the best moment to reconfigure solvers. Theoretical insights will be challenged through extensive empirical evaluations, ranging from configuring evolutionary algorithms for solving traditional black-box optimization problems to Bayesian Optimization for Hyperparameter Optimization of Deep Neural Networks.