Warmstarting of Model-based Algorithm Configuration
This addresses the tedious and suboptimal manual tuning of solver parameters for AI and combinatorial problem-solving, offering a practical improvement over existing AC methods.
The paper tackles the problem of algorithm configuration (AC) for combinatorial solvers by proposing warmstarting methods that use performance data from previous benchmarks to speed up configuration on new ones, resulting in up to 165-fold speedups and better configurations with the same compute budget.
The performance of many hard combinatorial problem solvers depends strongly on their parameter settings, and since manual parameter tuning is both tedious and suboptimal the AI community has recently developed several algorithm configuration (AC) methods to automatically address this problem. While all existing AC methods start the configuration process of an algorithm A from scratch for each new type of benchmark instances, here we propose to exploit information about A's performance on previous benchmarks in order to warmstart its configuration on new types of benchmarks. We introduce two complementary ways in which we can exploit this information to warmstart AC methods based on a predictive model. Experiments for optimizing a very flexible modern SAT solver on twelve different instance sets show that our methods often yield substantial speedups over existing AC methods (up to 165-fold) and can also find substantially better configurations given the same compute budget.