Ensemble Learning with Active Data Selection for Semi-Supervised Pattern Classification

Abstract

Unliketraditional pattern classification, semi-supervised learning provides a novel technique to make use of both labeled and unlabeled data for improving the performance of classification. In general, there are two critical issues for semi-supervised learning of discriminative classifiers; i.e., how to create an initial classifier of a good generalization capability with the limited labeled data and the how to make an effective use of unlabeled data without degradation of the established classifier. To tackle two aforementioned problems, we propose an ensemble learning approach based on a recent active data selection strategy [1], where ensemble learning would yield good generalization and active data selection tends to choose the unlabeled data more likely resulting in an improvement during semi-supervised learning. By using an ensemble of K-NN classifiers, we demonstrate the effectiveness of our approach on a synthetic data classification and a facial expression recognition tasks.

Time Series Clustering Via RPCL Network Ensemble With Different Representations

“…With such diversity, it is likely to reach the synergy to capture the intrinsic structure of a time series dataset. Inspired by our previous work in supervised and semisupervised ensemble learning [22]- [25], we come up with a model by the use of different representations for time series clustering. Fig.…”

“…Unlike the previous method [20], our approach is motivated by our previous success in the use of different representations to construct an ensemble model for dealing with difficult supervised [22]- [24] and semisupervised learning tasks [25], where the use of different representations better exploits the information conveyed in the raw data and therefore leads to the better performance. For each individual representation, we first employ an RPCL network for clustering analysis of automatic model selection, and the nature of an RPCL network often leads to quick clustering analysis.…”