In most data-mining applications where induction is used as the primary too
l for knowledge extraction from real-world databases, it is difficult to pr
ecisely identify a complete set of relevant attributes. This paper introduc
es a new rule induction algorithm called Rule Induction Two In One (RITIO),
which eliminates attributes in the order of decreasing irrelevancy. Like I
D3-like decision tree construction algorithms, RITIO makes use of the entro
py measure as a means of constraining the hypothesis search space; but, unl
ike ID3-like algorithms, the hypotheses language is the rule structure and
RITIO generates rules without constructing decision trees. The final concep
t description produced by RITIO is shown to be largely based on only the mo
st relevant attributes. Experimental results confirm that, even on noisy, i
ndustrial databases, RITIO achieves high levels of predictive accuracy.