Abstract
Classification is one of the major constituents of the data-mining toolkit. The
well-known methods for classification are built on either the principle of
logic or statistical/mathematical reasoning for classification. In this article we
propose: (1) a different strategy, which is based on the portioning of
information space; and (2) use of the genetic algorithm to solve
combinatorial problems for classification. In particular, we will implement
our methodology to solve complex classification problems and compare the
performance of our classifier with other well-known methods (SVM, KNN,
and ANN). The results of this study suggest that our proposed methodology
is specialized to deal with the classification problem of highly imbalanced
classes with significant overlap.
well-known methods for classification are built on either the principle of
logic or statistical/mathematical reasoning for classification. In this article we
propose: (1) a different strategy, which is based on the portioning of
information space; and (2) use of the genetic algorithm to solve
combinatorial problems for classification. In particular, we will implement
our methodology to solve complex classification problems and compare the
performance of our classifier with other well-known methods (SVM, KNN,
and ANN). The results of this study suggest that our proposed methodology
is specialized to deal with the classification problem of highly imbalanced
classes with significant overlap.
Originalsprog | Engelsk |
---|---|
Udgiver | Cambridge University Network |
Antal sider | 10 |
Status | Udgivet - nov. 2016 |