LBEF RESEARCH JOURNAL OF SCIENCE, TECHNOLOGY AND MANAGEMENT

Author(s):Jiten Limbu, Dr. Swati Sah

Abstract:More and more researchers are doing research on prediction model to predict student academic performance while dealing with large scale of educational databases. The current research and their products have not been able to give maximum accuracy due to various reasons like lack of appropriate data and lack of best algorithm. If existing systems can predict student’s academic progress and performance with 100% accuracy then further research on this field would not have been required. But to develop the prediction model on student’s academic performance with maximum accuracy rate is a very challenging work. These types of problems encourage the researcher to research on this field with data mining techniques. The application of data mining technique is done in order to divide information into separate clusters or classifications so that particular dataset can be studied under a classified group. The output of this prediction model helps to identify student’s educational status as well as his/her future performances by analyzing their data. It also helps to identify particular institution’s weak students and this prediction model can provide him/her with solutions by providing relevant suggestions to avoid the future failure. It will directly boost the performance of those identified weak students. So, educational institutions can easily take benefits by using this prediction model. The educational institutions can provide special package to improve those weak identified students by analyzing the student academic performance predictive model.

The main aim of this research is to develop student academic performance prediction model for the Nepalese bachelor level student in Computer Science stream using unsupervised learning with some data clustering method. Here, the Clustering Algorithm is the cluster-based unsupervised learning algorithm which clusters the particular dataset by analyzing its similarity. Our prediction model will be done through python programming language. One of the clustering algorithms will be implemented through this python programming language environment.

To conduct this research, each student’s internal subject marks are selected as parameter. The main objective of this research is to provide more description of the Clustering Algorithm and provide how the clustering algorithm will cluster or separate student’s academic internal marks and then develop the platform based on the research. A desktop-based application will be designed to get prediction model which predicts the student’s academic performance by implementing one of the clustering algorithms. The application will show the future performances of each particular student. It is also helpful to find out different categories of the student (i.e. weak, normal, intelligent) by analyzing their internal marks.

Keywords:Image Recognition, Plant Identification, Machine learning (ML), Deep Learning (DL), Neural Networks (NN), Convolutional neural network (CNN), Greyscale