Lift-based search for significant dependencies in dense data sets

Abstract

Dependency analysis is an important but computationally demanding problem in all empirical science. It is especially problematic in bioinformatics, where data sets are often high dimensional, dense and/or strongly correlated.As a solution, we introduce a new algorithm which searches the most significant association rules expressing positive dependencies. The algorithm uses several effective pruning principles, which enable search without any minimum frequency thresholds. According to our initial experiments, the algorithm suits especially well for typical biological and medical data sets.