Research And Application On The Evaluation System Of Student's Cognitive Ability In The NTS

The problem of student's cognitive ability's evaluation has been studied and concerned widely by the educational scholars. The problem is a more issue, it involves in the various aspects of morality, wisdom and health of students.It is a typical multiobjective multilayer comprehensive evaluation problem. Now, for the student's cognitive ability's evaluation, it has been many methods of evaluation .But the most are subjective, they contained many artificial factors. So, it is essential to seek a more reasonable, scientific comprehensive evaluation method.Based on the fuzzy theory and artificial neural network, this thesis researches and analyses the student's cognitive ability's evaluation based on fuzzy comprehensive evaluation model, BP neural network evaluation model, and RBF neural network evaluation model.Fuzzy comprehensive evaluation method is an evaluation method used for the object system involved fuzzy factors. So, using this method can solve the fuzzy problem of comprehensive evaluation problem effectively. And artificial neural network technology has highly self-adaptability, self-learning and self-organization. In the complicated system, the ANN shows to superiority in setting up model. So, it can properly be used to solve the comprehensive evaluation problem of complex, too. Both the BP neural network and RBF neural network are the feedforward type neural network, only the slightly different in the constructing methods and training methods. Thus, their training result is different.The main studying contents of this paper are as follows: (1)It establishes the target system of students' cognitive evaluation .(2)It establishes the Fuzzy Comprehensive Evaluation Model, and make a discussion on some critical problems of the Comprehensive Evaluation Model;and evaluates the students' cognitive ability using the two-stage comprehensive evaluation .(3) It sets up the BP neural network model, and analyses the various training parameters in the BP neural network ,implements it by MATLAB. (4)It establishes the RBF neural network model, and describes the Orthogonal Least Squares in it, ascertained the radical function'scenter. (5)Through researching and analyzing the above three models, we knew the RBF neural network more than other two evaluation models in the efficiency and accuracy. (6)It designs a students' cognitive ability evaluation system by using the fuzzy theory and artificial neural network with the program language of .NET(C#), SQL Server database and MATLAB toolbox.This paper expands the target of students' cognitive ability based on the educational scholars' researching result, let it accord with the educating goal. This evaluation system basically can achieve the evaluation for the students' cognitive ability in the morality,wisdom and health of aspects.