7月30日:Tadashi Dohi
发布时间:2025-07-25 浏览量:10

报告名称:Software Reliability Growth Model Averaging: Theory and Empirical  Validation

报告时间:7月30日15:00

报告地点滴水湖国际软件学院502

报告摘要:

Accurate prediction of the cumulative number of software faults  as well as quantitative software reliability has been still a challenging issue  in software reliability for a long time. Especially, weighted linear  combination models based on multiple software reliability growth models (SRGMs) have received a considerable attention to improve the prediction  accuracy, but have not been satisfactorily formulated during the last five  decades. In this talk, we develop several approaches to predict the future behavior  of the cumulative number of software faults detected in the system testing,  by means of multi-model inference, consisting of both the model averaging techniques, where the underlying SRGMs are described by the common non-homogeneous Poisson processes. The model averaging techniques employed  here possess an information-theoretic and frequentist perspective and are  characterized by well-known information criteria to determine each model  weight. Several clustering techniques are also applied to cluster theunderlying SRGM candidates for prediction We also improve the existing ensemble-based methods to predict the cumulative number of software faults in the similar framework to the SRGM averaging. In comprehensive numerical experiments with both time-domain and  time-interval software fault-count data, we compare our SRGM averaging  techniques with the common approach based on the single use of SRGM. It  is shown with actual software fault-count data that SRGM averaging enabled  us to improve prediction accuracy significantly.


报告人简介:

Tadashi Dohi received the Bachelor of Science in Engineering (B.S.E.), Master of Engineering (M.E.), and Doctor of Engineering (Dr. of Eng.)  degrees from Hiroshima University, Japan, in 1989, 1991, and 1995,  respectively. In 1992 he joined as an Assistant Professor in the Department  of Industrial and Systems Engineering, Hiroshima University. In 1992 and  2000, he was a Visiting Research Scholar in the Faculty of Commerce and  Business Administration, University of British Columbia, Canada, and  Hudson School of Engineering, Duke University, USA, respectively, on the leave absent from Hiroshima University. Since 2002, he has been  working as a Full Professor in Hiroshima University. He is now the Dean of  the School of Informatics and Data Science, and the Associate Dean of the  Graduate School of Advanced Science and Engineering, Hiroshima  University. His research areas include Reliability Engineering, Software Reliability,  Dependable Computing, Performance Evaluation, Computer Security and  Operations Research. He is a Regular Member of ORSJ, IEICE, IPSJ, REAJ  and IEEE. He published over 644 peer-reviewed papers (285 journal papers  and 359 conference papers), 49 book chapters, and 26 edited books/proceedings in the above research area. Dr. Dohi served as a General  Chair/Program Committee Chair of over 20 international events and worked  as a Program Committee Member in many international premier conferences  such as DSN, ISSRE, EDCC, SRDS, HASE, QRS, ICPE, SAC, among  others. He is an Associate Editor/Editorial Board Member of a dozen of  international journals. He is one of Founders of AIWARM/APARM and  served as the Steering Committee Chair.


华东师范大学软件工程学院
学院地址:上海中山北路3663号理科大楼

                上海市浦东新区楠木路111号
院长信箱:yuanzhang@sei.ecnu.edu.cn | 办公邮箱:office@sei.ecnu.edu.cn | 院办电话:021-62232550
www.sei.ecnu.edu.cn Copyright Software Engineering Institute