Article Preview
Top1. Introduction
One of the crucial elements in our society is education. Education is a tool to change the status of an individual in society. In recent days, the educational institutions are maintaining students’ details such as their enrollment details, personal details, previous study details, attendance details, internal and external marks details in huge volumes. To analyze and predict the students’ performance, various Data Mining techniques are applied on the datasets. Educational Data Mining (EDM) is one of the interdisciplinary research fields where numerous studies have been carried out to develop the quality of education and for predicting the academic performance of students at earlier (Raheela Asif, 2017). Though, various techniques have been used to predict the student academic performance, till it is a challenging task. Because the performance of students depends on various factors such as personal, socio-economic, psychological, academic background, family background and on other environmental variables (Ashwin Satyanarayana, 2016).
Learners, learning processes and learning situations are the three important parts of education. The academic class is generally heterogeneous (i.e.) Fast, Average and Slow learners are there in each class. The responsibilities of educators in earlier education systems were limited only with educating the students in the classroom to increase the knowledge of students. In current days, the educators should focus on the overall development of the students. Therefore, the responsibilities of each academic institution are to provide appropriate direction to the students’ for selecting the right carrier based on their abilities, skills and interest. Appropriate directions can help the students to attain their learning goals and also to acquire satisfaction in their personal life. Academic Institutions have severe competition among one another, trying to attract the student who will successfully pass through the educational process and making efforts to handle with student retention. Also, the educational institutions are very often forced to take quick decisions, therefore timely and high-quality information of student is needed (Kabakchieva, 2013). The Educational Data Mining concentrates to explore the data originating from the educational environments by developing new models. In learning process, students are passing through several physical development stages where, each stage has its own features. If the respective teacher identifies these features then the teacher can assist the learners in imparting instruction and changing their activities as for the specified educational intention (Chia, 2017).
Machine Learning is the science of making computers to act without any human intervention or explicit programs. In Machine Learning, there are two types of classification. One is Binary classification and the other is Multiclass classification. In Binary classification, the machine should classify an instance either in one of two classes (i.e.) yes/no or 1/0 or true/false. But in Multiclass classification, an instance can be classified as only one of three classes or more. Predictive modeling is the branch of machine learning that deals with data to explicitly find patterns from the given data. Through this predictive modeling, we are predicting the student academic performance at earlier stage (Farshid Marbouti, 2016).
1.1 Objectives
- 1.
Identifying the key factors that correlate with students’ performance.
- 2.
Analyzing the importance of the factors and selecting the high dominant features that affect the academic performance of students.
- 3.
Developing models to classify the learners as fast, average or low performers.
1.2 Outline of the Paper
Organization of this work is as follows. The related works on student’s academic performance prediction were presented in Section 2. In Section 3, the data and research methods have been described. Section 4 describes the results and discussions. Conclusion is given at Section 5.