GuideR: a guided separate-and-conquer rule learning in classification, regression, and survival settings
This addresses the need for more interpretable and user-controlled rule learning in data analysis, though it appears incremental as it builds on sequential covering approaches.
The authors tackled the problem of existing rule induction methods lacking user guidance, which often leads to uninteresting rules, by proposing GuideR, a guided separate-and-conquer algorithm that incorporates user preferences and domain knowledge. The method was experimentally verified to be effective in classification, regression, and survival analysis tasks.
This article presents GuideR, a user-guided rule induction algorithm, which overcomes the largest limitation of the existing methods-the lack of the possibility to introduce user's preferences or domain knowledge to the rule learning process. Automatic selection of attributes and attribute ranges often leads to the situation in which resulting rules do not contain interesting information. We propose an induction algorithm which takes into account user's requirements. Our method uses the sequential covering approach and is suitable for classification, regression, and survival analysis problems. The effectiveness of the algorithm in all these tasks has been verified experimentally, confirming guided rule induction to be a powerful data analysis tool.