``Sampling Theories for Rule Discovery Based on Generality and Accuracy:
the Worst Case and a Distribution-Based Case'',
Communication of Institute of Information and Computing Machinery, Taiwan,
Vol. 5, No. 2, pp. 83-88, May 2002.
Fumio Takechi and Einoshin Suzuki:
``Finding an Optimal Gain-Ratio Subset-Split Test for a Set-Valued Attribute in Decision Tree Induction'',
Proc. Nineteenth International Conference on Machine Learning (ICML),
pp. 618-625, July 2002, Sydney.