Hillary is a general learning system that selectivly acquire macro-operators for speeding up search programs. Solvable training problems are generated in increasing order of difficulty. The only macros acquired are those that take the problem solver out of a local minimum to a better state. The utility of the method was demonstrated in several domains, including the domain of NxN sliding-tile puzzles. After learning on small puzzles, the system is able to efficiently solve puzzles of any size. Hillary was developed by Lev Finkelshtein and Shaul Markovitch.

Designed by Ofer Avnery & Eyal Fingold