Artificial Intelligence Applied to Assistive Technologies and Smart Environments, Papers from the 2016 AAAI Workshop, Phoenix, Arizona, USA, February 12, 2016

Automatic algorithm configuration addresses the problem of determining the settings of an algorithm’s parameters to optimize its performance. It has most prominently been applied to optimize solvers for hard combinatorial problems (for example, SAT, MIP, ASP, and AI planning) as well as to hyperparameter optimization of flexible machine-learning frameworks (such as deep neural networks, or the space of algorithms defined by the Waikato environment for knowledge analysis, WEKA), but it also has applications in many more areas. Additionally, algorithm configuration has been performed manually by domain experts in a tedious and time-consuming optimization process, a task humans are poorly suited for.