Oversearching and Layered Search in Empirical Learning

When learning classifiers, more extensive search for rules is shown to lead to lower predictive accuracy on many of the leal-world domains investigated. This counter-intuitive re suit is particularly relevant to recent system the search methods that use risk-free pruning to achieve the same outcome as exhaustive search. We propose an iterated search method that commences with greedy search extending its scope at each Iteration until a stopping criterion is satisfied. This layered search is often found to produce theories that are more accurate than those obtained with either greedy search or moderately, extensive beam search.