An enhancement dbscan algorithm applied to faculty performance evaluation

ABSTRACT: DBSCAN is a fundamental density cluster formation. Its advantage is that it can discover clusters with arbitrary shapes and size. The algorithm typically regards clusters as dense regions of objects in the data space that are separated by regions of low-density objects. The algorithm has two input parameters, radius Eps and MinPts. It only needs to find out all the maximal density connected spaces to cluster the data points in an attribute space. And these density-connected spaces are the clusters. Every object not contained in any cluster is considered noise and can be ignored. Though DBSCAN has a lot of advantages compared to other clustering algorithms, it also has its own drawbacks. First, the algorithm is not entirely deterministic wherein it cannot determine the correct cluster of a border point. Second, DBSCAN is sensitive to the setting of parameters resulting to unfitting clusters. Third, possible merging of two supposedly separate clusters may occur in the existing algorithm, failing to comply with the goal of cluster analysis. The researchers were able to enhance the existing system by solving the problems stated and have met the objectives presented. The enhanced algorithm satisfied the standards of the possible user through the conducted survey.

Filipiniana