No abstract

A machine leaming technique called Graph-Based Induction (CBI) extracts pattems from graph structured data by stepwise pair expansion. GEI has been extended to I) Beam-wise GBI (B-GEI} by incorporating a beam search to improve its search capabilig, and 2) Decision Tree GraphBased Induction (DT-CBI) to construct a decision tree for graph-structured data. We applied B-CBI and DT-CBI to analyze the eflectiveness of inteqeferon therapy in the hepatitis dataset provided by Chiba University Hospital. Descriptive patterns were extracted by B-GBI and discriminative ones by DT-GBI using only the time sequence data of blood inspection and urinalysis. The discriminative patterns e.rtracted by DT-GBI tend to be included in only relatively small number of patients and thus too specific. Thus, we tried to extract patterns which are both discriminative and descriptive by B-GBI. Furthermore, since there are exceptional situations (patients) with the extracted patterns, these partems are further utilized tu extracf re$ned knowledge from the dataset. The preliminary results are reported with some uf extracted patterns.

No abstract