Supervised Learning to Control Energetic Reasoning: A Feasibility Study

Van Cauwelaert, Sascha;Lombardi, Michele;Schaus, Pierre
(2014)

Files

seer-dp.pdf
  • Open Access
  • Adobe PDF
  • 286.78 KB
Download

Details

Authors
  • Van Cauwelaert, SaschaUCLouvain
    Author
  • Lombardi, MicheleUCLouvain
    Author
  • Schaus, Pierreorcid-logoUCLouvain
    Author
Abstract
(en) Propagation is a double-edged sword, with more pruning power coming at the price of larger computation time. For each problem constraint, the best propagator depends on the specific instance and may change at search time. We propose to use an oracle function, obtained via Machine Learning, to decide whether to run complex propagators for a target constraint. In this paper, we focus on investigating the feasi- bility of building an oracle for the Energetic Reasoning propagator used in scheduling. Our experiments show that high prediction accuracy can be obtained, provide suggestions for classification features, and highlight important issues to address when building such an oracle.
Affiliations

Citations

  • APA
  • Chicago
  • FWB

Van Cauwelaert, S., Lombardi, M., & Schaus, P. (2014). Supervised Learning to Control Energetic Reasoning: A Feasibility Study. https://hdl.handle.net/2078.5/231514