Mind Evolutionary Computation with Lévy Search and Idea Recombination

Summary

The document explains Mind Evolutionary Computation (MEC), a population optimizer modeled on competition and information exchange among groups. In the basic method, agents form groups, search around each group’s best candidate, and use group comparisons to replace weak groups and introduce new ones. Message boards retain each group’s current best result and guide these operations.

The author then describes a variant that treats candidates as ideas with parameter theses. It uses Lévy-distributed branching for local exploration, exchanges strong ideas from an alternative group into a dominant group, and creates new candidates by combining theses from dominant ideas. The article says this recombination can improve search diversity and reports comparative algorithm tests, but the provided excerpt does not include enough detail to assess those results. It also lists getting stuck on flat regions as a weakness. The method is presented as a general optimization approach, not a trading strategy, and the author notes that the implementation includes personal modifications to the canonical algorithm.

Key ideas

• MEC searches through groups of candidate solutions using local competition and group-level replacement.
• Message boards preserve group winners and inform the next search steps.
• The described variant uses Lévy-distributed branching to explore around ideas.
• It recombines parameters from strong candidates to create new ideas and increase combinatorial search.
• The author identifies flat fitness regions as a potential source of stagnation.

Tags

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.