Peng Yijie's research group of Guanghua School of Management published a paper proposing an efficient simulation optimization method to solve the problem of personalized decision-making in large-scale complex stochastic systems

Recently, an article by Peng Yijie's research group at Peking University's Guanghua School of Management titled "Efficient Learning for Clustering and OptimizingContext-Dependent Designs" was accepted
by Operations Research 。 Operations Research has been recognized as the flagship journal
in operations research and management science for publishing theoretical methods such as stochastic modeling, simulation, optimization and application in the fields of energy environment, financial engineering, operations management, and medical services.

Screenshots of the journal

Simulation optimization is an active research field with a long history in operations research, with the aim of efficiently optimizing large-scale complex stochastic systems through simulation sampling, and traditional research problems do not consider personalized decision-making
in different scenarios.

Simulation model of Markov chain for esophageal cancer and its preventive treatment

In order to more effectively mine the random simulation sample information, the author introduces the Gaussian hybrid model to characterize the clustering phenomenon, and designs an efficient approximation method to reduce the computational complexity from exponential rate to linear rate, which solves the difficulties
in the theory and application of the random clustering method based on simulation.

Clustering and optimization results of personalized treatment options based on esophageal cancer preventive treatment simulation model

In addition, a recent article by Peng Yijie's research group entitled "Efficient learning for decomposing and optimizing random networks" was published
in Fundamental Research, a journal sponsored by the National Natural Science Foundation of China 。 In the background of the PageRank web page sorting method proposed by Google, the simulation optimization method of efficiently learning the clustering of random networks and sorting the importance is proposed by random sampling, and the asymptotic nature of the algorithm is theoretically analyzed, and the application in the real Internet sorting problem has achieved better performance
than the traditional method.

The first author, Li Haidong, is currently a postdoctoral fellow at the School of Engineering, Peking University, and received his bachelor's degree from the School of Engineering of Peking University in 2015 and his Ph.

Peng Yijie, the corresponding author of this article, is currently an associate professor in the Department of Management Science and Information Systems of Guanghua School of Management, Peking University, and a part-time researcher at
the Institute of Artificial Intelligence of Peking University and the National Institute of Health and Medical Big Data.