Improved classification EM algorithm for the problem of separating semiconductor device production batches

International Workshop "Advanced Technologies in Material Science, Mechanical and Automation Engineering – MIP: Engineering – 2019"

材料科学;机械制造;原子能学

Rozhnov, I.^1 ; Kazakovtsev, L.^1^2 ; Bezhitskaya, E.^1 ; Bezhitskiy, S.^1^2

Reshetnev Siberian State University of Science and Technology, prosp. Krasnoyarskiy Rabochiy 31, Krasnoyarsk

660031, Russia^1

Siberian Federal University, 79 Svobodny av., Krasnoyarsk

660041, Russia^2

关键词: CEM algorithms;    Classification EM;    Cm algorithms;    Greedy heuristics;    Numerical experiments;    Objective functions;    Production batches;    Standard deviation;   

Others  :  https://iopscience.iop.org/article/10.1088/1757-899X/537/5/052032/pdf
DOI  :  10.1088/1757-899X/537/5/052032

学科分类：材料科学（综合）

来源: IOP

【 摘 要 】

This paper focuses on new proposed algorithms for cluster problem solving. The proposed algorithms are based on Classification EM algorithm (CM-algorithm). The algorithms are new algorithms of the greedy heuristic method using the idea of searching in alternating neighborhoods. The numerical experiments show that the proposed algorithms have less mean values and/or less standard deviation of objective function, less scatter of obtained values in comparison with classical CEM-algorithm.