抽象的

Intellectual Performance Analysis of Students by Using Data Mining Techniques

J.K. Jothi Kalpana, K. Venkatalakshmi

Education Data Mining concerns the prediction of school failures in different levels such as primary, secondary and higher level. This paper intends to analysis the students’ performance in different categories of measurements. In this analysis categorize the college student’s academic performance for Villupuram district. Based on the clustering methods such as centroid based, distribution based and density based clustering. Cluster includes groups with small distance among the cluster members. The performance of student’s multi-level of optimization formulated by using clustering. In centroid based clustering, clusters are represented by a central vector. The number of clusters is fixed to k, k-means clustering gives a formal definition as an optimization problem. The clustering model most closely related to statistics is based on distribution model. Experiments attempts to improve the accuracy by using the method of Gaussian mixture model. The data set is modeled with a fixed number of Gaussian distribution that is initialized randomly and the parameters are iteratively optimized to fit better to the data set. The density based clustering method is a linkage based clustering. The range parameter ε produces a hierarchical result related to that of linkage clustering. Clustering can be represents in a large range of classifications and applications. K-means algorithm categorizes the large dataset. In this analysis use genetically improved particle swarm optimization algorithm to model the students level. The GAI-PSO algorithm searches the solution space to find the optimal result. The processing of refining use the k-means algorithm.