Classification of Uncertain Data Streams Based on Extreme Learning Machine

Abstract

Classification over data streams is an important task in data mining. The challenges become even larger when uncertain data are considered. An important challenge in the classification of uncertain data streams is concept drift and uncertainty of data. This paper studies the problem using extreme learning machine (ELM). We first propose weighted ensemble classifier based on ELM (WEC-ELM) algorithm, which can dynamically adjust classifier and the weight of training uncertain data to solve the problem of concept drift. Furthermore, an uncertainty classifier based on ELM (UC-ELM) algorithm is designed for the classification of uncertain data streams, which not only considers tuple value, but also its uncertainty, improving the efficiency and accuracy. Finally, the performance of our methods is verified through a large number of simulation experiments. The experimental results show that our methods are effective ways to solve the problem of classification of uncertain data streams and are able to solve the problem of concept drift, reduce the execution time and improve the efficiency.