Grafting for Combinatorial Boolean Model using Frequent Itemset Mining
This work addresses the computational bottleneck in learning interpretable Boolean models for data analysis, representing an incremental improvement with a novel method for a known bottleneck.
The paper tackles the problem of learning combinatorial Boolean models (CBM) from labeled data, which are interpretable but computationally expensive to learn naïvely, and proposes the GRAB algorithm that reduces this to weighted frequent itemset mining, showing effectiveness in computational efficiency, prediction accuracy, and knowledge discovery on benchmark datasets.
This paper introduces the combinatorial Boolean model (CBM), which is defined as the class of linear combinations of conjunctions of Boolean attributes. This paper addresses the issue of learning CBM from labeled data. CBM is of high knowledge interpretability but naïve learning of it requires exponentially large computation time with respect to data dimension and sample size. To overcome this computational difficulty, we propose an algorithm GRAB (GRAfting for Boolean datasets), which efficiently learns CBM within the $L_1$-regularized loss minimization framework. The key idea of GRAB is to reduce the loss minimization problem to the weighted frequent itemset mining, in which frequent patterns are efficiently computable. We employ benchmark datasets to empirically demonstrate that GRAB is effective in terms of computational efficiency, prediction accuracy and knowledge discovery.