Wang, Chao; Guo, Jing; Bao, Zhen-qiang Application of improved Q learning algorithm to job shop problem. (Chinese. English summary) Zbl 1171.68684 J. Comput. Appl. 28, No. 12, 3268-3270 (2008). Summary: The Job Shop Problem (JSP) is a key technology in manufacturing system, and the Q learning is used to realize it. The improved Q learning algorithm is suggested because of the traditional algorithm has limitations of slow and partial constringency. In this algorithm, a complex action group is suggested. Uniformity degrees of action choices are weighted by cohesion, and then the two limitations can be overcome effectively. It can be adapted to the complicated manufacturing environment. Experimental results show its good effect in the JSP. MSC: 68T05 Learning and adaptive systems in artificial intelligence Keywords:Job Shop Problem (JSP); reinforcement learning; Q learning; cohesion PDFBibTeX XMLCite \textit{C. Wang} et al., J. Comput. Appl. 28, No. 12, 3268--3270 (2008; Zbl 1171.68684) Full Text: DOI