Research on Uncertain Prediction Method Based on Credibility Distribution

• Published in: Learning Technologies and Systems Vol. 12511; pp. 148 - 159
• Main Authors: Li, Yan, Jin, Chenxia, Wang, Ying
• Format: Book Chapter
• Language: English
• Published: Switzerland Springer International Publishing AG 2021; Springer International Publishing
• Series: Lecture Notes in Computer Science
• Subjects: Big data; Credibility distribution; Prediction; Sampling with replacement
• ISBN: 303066905X; 9783030669058
• ISSN: 0302-9743; 1611-3349
• DOI: 10.1007/978-3-030-66906-5_14

With the rapid development of information society, traditional data mining is unable to meet the actual needs. In this paper, an uncertain prediction method based on credibility distribution (RDP) is proposed. Firstly, the implementation mechanism of the uncertain prediction based on credibility distribution in sampling is given. Secondly, combining with the law of large Numbers, the convergence characteristics of test credibility of the decision attribute corresponding to the value of a conditional attribute in sampling are analyzed. Finally, the validity of RDP is verified through Simulation experiment of UCI database. Theoretical analysis and simulation results show that RDP is feasible in interpretability and operability.