The Mining Algorithm Of Classification In The Teacher Model

In the system of Intellectual Computer-assisted Instruction(ICAI) ,teaching students in accordance with their ability is the main embodiment of intelligent, rational classification of students is the base of it and can provide different teaching tactics in accordance with different types of students. The important problem that should be resolved in the model of teacher is the classification of students and then the model has the individualized characteristic,this is necessary to a good model of teacher. Meamwhile,it has the auxiliary function in a teacher's practice. So,constructing a fast , effective categorised algorithm is the base of the achievement of individualized teacheng, it is important to the realization of intelligent in the ICAI. In this paper,Cluster analysis and principal components in the data mining are be used to make a series of describing mining on the scores of students and set up the classification of students. At first, there are some relation among every classes,or the college entrance examination is inconsistent with the course offered of the university, so it is difficult to make sure the ability from the scores.Before the classification of students,principal components is used to carry on the pretreatment on the scores and then some new irrelative indexes are used to replace the already existing indexes,it is aimed at rational classification. The question of finding characteristic values and characteristic vectors of a real symmetrical matrix is involved. The Jacabi algorithm is used because of the probability of parallel computation in the future. Then cluster analysis is used to classify the students according to the new indexes after the pretreatment in the paper.If the scores of many times are availuable, the M-measure can be used in the cluser analysis,and then the factor of relation among the indexes is considered in this measure. But the problem of finding the inverse of a symmetrical positive matrix is involved by the computition of M-measure betweem the cases. When many indexes , adopting the general algorithm will led to the high complexity of calculation. The newer rotation algorithm is used to calculate the M-measure in this paper,and this has improved the computational speed greatly . Meanwhile, the structure of highly education in our country is changing greatly at entrance of 21st century, students are given large freedom to choose the beloved and suitable speciality by theirselves.At present, some universities have implemented the uncertain professionaldirection in the first year and confirm the professional direction in the second year. so students have sufficient time to decide professional direction in the future. This is integrated with the foreign teaching system and is the main trend of development of educational structure.Students' choice of professional direction is very cautious, it most likely determines the job that a person is engaged in all his life. Generally speaking, whether one can make progress at study is mainly depend on two points. Firstly, the interest , hobby and personal characteristic, secondly, his trend of ability. Different ability is suitable for different professional directions. In the model of teacher, When consulting or instructing every student, we can carry on a scientific classification according to students' college entrance examination achievement ,and divide students into several groups according to a certain ability ,and analyse the aptiudes of each kinds according to the character of every course, then offer scientific guidance in professional choice of students. So this paper has strong practicability.