Research on an advanced intelligence implementation system for engineering process in industrial field under big data
Yuan Guo,Bing Zhang,S. Yu,W. Kai
DOI: https://doi.org/10.1016/j.eswa.2020.113751
IF: 8.5
2020-12-01
Expert Systems with Applications
Abstract:<p>To develop an advanced CBR system to well adapt to the intelligence implementation of new engineering process in the big data environment, Bayesian network (BN) model is introduced to CBR system for knowledge reasoning. However, as engineering application is becoming more and more complicated, the number of parameters used to define engineering application grows larger and larger, leading to the seriously reduced efficiency as well as the accuracy of the integrated model. For the problem of reduced efficiency, this paper proposes In-External (IE) algorithm to perform the assignment of big data distribution for parallel data processing, which can fully utilize the capacity of Hadoop system and attain the best efficiency of knowledge reasoning. For the problem of reduced accuracy, in view of the fact that traditional probability learning methods are unfit for the proposed CBR system, this paper proposes Discount Exponential Coefficients of Multivariate Beta Distribution (DECMBD) algorithm to conduct the probability learning of proposed system. In DECMBD algorithm, a discount ratio is given to each exponential coefficient of multivariate Beta distribution to improve the occurrence times counting of all problem features and then gain better effect of probability learning. Finally, lots of experiments are performed to validate the effectiveness of the proposed advanced CBR system.</p>
computer science, artificial intelligence,engineering, electrical & electronic,operations research & management science