Universal Search is an asymptotically optimal way of searching the space of programs computing solution candidates for quickly verifiable problems. Despite the algorithm's simplicity and remarkable theoretical properties, a potentially huge constant slowdown factor has kept it from being used much in practice. Here we greatly bias the search with domain-knowledge, essentially by assigning short codes to programs consisting of few but powerful domain-specific instructions. This greatly reduces the slowdown factor and makes the method practically useful. We also show that this approach, when encoding random seeds, can significantly reduce the expected search time of stochastic domain-specific algorithms. We further present a concrete study where Practical Universal Search (PUnS) is successfully used to combine algorithms for solving satisfiability problems.
CITATION STYLE
Schaul, T., & Schmidhuber, J. (2010). Towards practical universal search. In Artificial General Intelligence - Proceedings of the Third Conference on Artificial General Intelligence, AGI 2010 (pp. 139–144). Atlantis Press. https://doi.org/10.2991/agi.2010.19
Mendeley helps you to discover research relevant for your work.