Collaborative learning for improving intellectual skills of dropout students using datamining techniques

In each year millions of students drop out without
completing their educational course. In such a case, both the individual student and institution will have an effect of dropping out. The proposed research work pays significant
research attention towards analysing the higher education and upper primary students to identify their behaviour, which leads them to discontinue in the early stage and stop the dropout by taking necessary action towards the dropout
reason. This in turn results in the lack of skilled workspace and weaken the productive system of the country and also student dropouts are more likely to become as the recipients of unemployment subsidies. This research is more focused on the dimension reduction techniques, which involves both the feature selection and feature extraction methods. It also aims to implement better prediction in identifying dropout students and implement collaborative learning with engagement. When PCR measures, the key elements do not look at the reaction but rather at the predictors (by looking for a linear combination of the predictors that has the highest variance). It assumes that, the answer is correlated with the linear combination of the predictors with greatest variance. It is presume that, the regression plane differs when selecting the main variable in the other orthogonal direction, along the line and it does not differ. The second path is disregarded by selecting one component and not the other. Principal Component Analysis (PCA) is a method used for extracting features that database. This is illustrated in two phases. First phase is the development of dimension reduction using PCA to identify an accurate prediction variance of dropout students by using
various ML algorithms and the second phase involves the
developing of collaborative learning with engagement through
social media and improves their intellectual skills by
performing SVM hypothesis test.