Ensemble Classifiers for DNA Microarray Data Analysis(Biometrics1)

概要

論文の詳細を見る
The development of microarray technology has provided a great amount of data to many areas. In particular, it has helped to predict and diagnose cancer. Because the problem of diagnosing and predicting cancer has been an important issue, many machine learning techniques has been studied and applied to produce informative results. These classification methods have some practical limitations, because microarray data can be noisy and incomplete, and classification algorithm itself cannot be perfect. This paper presents three sophisticated methods to solve these problems, which exploit the ensemble approach in common. Using multiple features of data and combining the results of multiple classifiers, more accurate prediction can be obtained. Experiments with lymphoma and colon cancer datasets have shown the usefulness of our proposed methods.
社団法人電子情報通信学会の論文
2006-11-17