Professor Shulan Hu and Associate Professor Xinyu Wang, the researchers of the base, has published a collaborative paper titled "Learning Performance of the Perceptron Model Based on Markov Sampling" in Acta Mathematicae Applicatae Sinica.

Abstract: The perceptron model is an essential algorithm in machine learning andartificial intelligence, functioning as a linear classifier for binary classification tasks.This paper investigates the learning performance of the perceptron model based onMarkov sampling, which builds upon the traditional framework with independent andidentically distributed samples. Initial efforts establish the constraints on the percep-tron model’s learning performance with uniformly ergodic Markov chain samples andvalidates its consistent behavior. Furthermore, the paper introduces a ueMC-PM algo-rithm. Numerical investigations undertaken on benchmark repositories reveal that theperceptron model utilizing ueMC samples yields fewer misclassification rates.

Keywords: learning performance; Markov sampling;perceptron model; uniformly ergodic Markov chain

Link: 基于马尔可夫抽样的感知机模型的学习性能分析

Teacher profile

Shulan Hu, Professor, Doctoral Supervisor, Wenlan Young Scholar, Young Teacher "Research Star", School of Statistics and Mathematics, Zhongnan University of Economics and Law. Director of the Resource and Environment Branch and the Big Data Branch of the Chinese Association for Applied Statistics, and Director of the Hubei Association for Applied Statistics. She mainly engages in research on theoretical and applied research on big data statistical algorithms, econometric methods and their financial applications, and other aspects. She has published more than twenty papers in domestic and international authoritative academic journals such as Bernoulli, Statistics Sinica, Stochastic Processes and their Applications, Science in China, and Zhongguo Kexue (Chinese Science). She has independently authored one academic monograph and served as the chief editor of the "'14th Five-Year' National Statistical Planning Textbook Econometrics". She has hosted and completed more than ten projects, including the Youth Project of the National Natural Science Foundation of China, the General Project of the National Social Science Foundation, central university research projects, graduate excellent course construction, all-English course construction projects, and horizontal projects for enterprises and public institutions. She has been awarded first and second prizes for guiding students in various competitions such as the National Market Research Competition, National Statistical Case Competition, National Industrial and Economic-Financial Big Data Modeling and Computing Competition, and American Mathematical Modeling Competition.

Xinyu Wang, Associate Professor, Wenlan College, Zhongnan University of Economics and Law; Ph.D. in Applied Mathematics, University of Côte d'Azur (UCA), France. Research Interests: big data statistical analysis, stochastic algorithm theory, applications of stochastic processes and stochastic analysis in economics and finance.