心理软计算的理论和应用  

The Theory and Application of Psychological Soft Computing

作　　者：余嘉元[1] 

机构地区：[1]南京师范大学心理学院,南京210097

出　　处：《心理科学》2016年第4期770-776,共7页Journal of Psychological Science

摘　　要：心理软计算是软计算、机器学习、数据挖掘、智能计算和心理学相结合的新方法体系,本文提出了心理软计算的总体框架和基本体系,指出该理论的主要内容包括心理软测量、心理软识别、心理软建模和心理软评估,介绍了它在参数估计、认知诊断、数据分析、综合评估和测验编制等方面的应用。随着现代科学技术的发展,心理软计算的内涵必将更加丰富,应用一定更加广泛。Psychological soft computing is a new method/system that combines soft computing, machine learning, data mining, intelligent computing and psychology. Soft computing includes artificial neural network, fuzzy logic, simulated annealing algorithm, rough set, support vector machine, genetic algorithm, ant colony algorithm, particle swarm optimization, gene expression programming and other swarm intelligence algorithms. It can deal with uncertain and unclear data in psychology, and obtain valuable information. In this paper, the general framework and basic system of psychological soft computing are presented. The main contents of this theory include psychological soft measurement, soft identification, soft modeling and soft assessment. Psychological soft measurement is indirect measurement. It uses the subjects＇ behavior information, which is easily available, and estimates its latent trait parameters using the modeling and calculating methods. The core of soft measurement is the establishment of the model, and the model is used to carry out the parameter inversion. Psychological soft measurement breaks through the framework of traditional statistical models. It uses the neural network and supports vector machine methods to carry on the parameter inversion. Psychological soft recognition is a classification with uncertainty. It includes feature selection, extraction and intelligent recognition. Cognitive diagnosis, as the current research focus, is essentially soft recognition. Soft computing methods of feature selection and extraction are the neural network, rough set and genetic algorithm. The methods used most frequently in intelligent pattern recognition are neural network, fuzzy set, support vector machine. In recent years, deep learning has created a new era of neural network development. It greatly improves the recognition accuracy. It is not only used in speech and image recognition, but also in the field of advanced cognitive activities of human beings. Psychological soft modeling is to set up the rela

关 键 词：心理软计算 软测量 软识别 软建模 软评估 

分 类 号：B841[哲学宗教—基础心理学] TP18[哲学宗教—心理学]