Tri-training: exploiting unlabeled data using three classifiers

• Published in: IEEE transactions on knowledge and data engineering Vol. 17; no. 11; pp. 1529 - 1541
• Main Authors: ZHOU, Zhi-Hua, MING LI
• Format: Journal Article
• Language: English
• Published: New York, NY IEEE 01-11-2005; IEEE Computer Society; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Algorithms; Applied sciences; Artificial intelligence; Classification; Classifiers; co-training; Computer science; control theory; systems; Data mining; Exact sciences and technology; Humans; Index Terms- Data mining; Information systems. Data bases; Labeling; Learning; Learning and adaptive systems; learning from unlabeled data; Machine learning; Machine learning algorithms; Memory organisation. Data processing; Parameter estimation; Partitioning algorithms; semi-supervised learning; Semisupervised learning; Software; Supervised learning; Tasks; tri-training; Web page classification; Web pages; Websites; World Wide Web; Data analysis; Redundancy; Information extraction; Data mining; machine learning; Index Terms-Data mining; tri-training; Supervised learning; Classification; World wide web; Internet; co-training; semi-supervised learning; Learning algorithm; learning from unlabeled data; Web page classification
• ISSN: 1041-4347; 1558-2191
• DOI: 10.1109/TKDE.2005.186

In many practical data mining applications, such as Web page classification, unlabeled training examples are readily available, but labeled ones are fairly expensive to obtain. Therefore, semi-supervised learning algorithms such as co-training have attracted much attention. In this paper, a new co-training style semi-supervised learning algorithm, named tri-training, is proposed. This algorithm generates three classifiers from the original labeled example set. These classifiers are then refined using unlabeled examples in the tri-training process. In detail, in each round of tri-training, an unlabeled example is labeled for a classifier if the other two classifiers agree on the labeling, under certain conditions. Since tri-training neither requires the instance space to be described with sufficient and redundant views nor does it put any constraints on the supervised learning algorithm, its applicability is broader than that of previous co-training style algorithms. Experiments on UCI data sets and application to the Web page classification task indicate that tri-training can effectively exploit unlabeled data to enhance the learning performance.