A Novel Extreme Learning Machine-Based Classification Algorithm for Uncertain Data

Abstract

Traditional classification algorithms are widely used on determinate data. However, uncertain data is ubiquitous in many real applications, which poses a great challenge to traditional classification algorithms. Extreme learning machine (ELM) is a traditional and powerful classification algorithm. However, existing ELM-based uncertain data classification algorithms can not deal with data uncertainty well. In this paper, we propose a novel ELM-based uncertain data classification algorithm, called UELM. UELM firstly employs exact probability density function (PDF) instead of expected values or sample points to model uncertain data, thus avoiding the loss of uncertain information (probability distribution information of uncertain data). Furthermore, UELM redesigns the traditional ELM algorithm by modifying the received content of input layer and the activation function of hidden layer, thus making the ELM algorithm more applicable to uncertain data. Extensive experimental results on different datasets show that our proposed UELM algorithm outperforms the baselines in accuracy and efficiency.