In most of the papers written on ID3 and C4.5 classifiers the example taken is the weather dataset of 14 days related with the playing of ‘Base-Ball’. The authors have taken these 14 samples for training the system of the classifier .We will show in this paper that only 5 samples are sufficient for training the classifier with the four weather-attributes viz. Outlook, Temperature, Humidity & Wind-speed. Actually, there will be a total of 36 combinations with these 4 attributes with 3 level values in Outlook (namely Sunny, Overcast &Rain), 3 level values in Temperature (namely Hot, Mild & Cool), 2 level values in Humidity (namely High &Normal) and 2 level values in Wind-Speed (namely Strong & Weak). (i.e. 3x3x2x2=36 combinations) .The sufficiency requirement of 5 samples is proved from Multi-level Boolean functions of switching theory.

Full text: IJEIR_588_Final.pdf