An Application that Predicts Grades of Students using a Simple Linear Regression Model and an Extremely Ideal Dataset
Abstract
The objective was to build a website that will help teachers understand the relationship between different subjects and extracurricular activities. We wanted to present an easy user interface using which, the teachers could add data regarding the student, search for data regarding a specific student and recognize the correlation between different subjects. For example, if a student is good at Chemistry, is he likely to be good at Biology as well? This is what the website helps the teacher understand. The dataset we used has data regarding ~500 students and their scores in Math, Chemistry, Computer Science, Biology and their interest in Sports.
Ajax Pattern: Submission Throttling
Using Submission Throttling, you buffer the data to be sent to the server on the client and then send the data when the user is idle for more than a stipulated amount of time. It doesn't send a request after each character is typed. We have used this to minimize the request while the teacher searches for a student’s information.
Data Science Component: Linear and Logistic Regression
-
We are using Linear Regression to predict the scores of subjects based on the scores of the subjects that they have a high correlation with. For instance, in our dataset, we observe that Mathematics and Computer Science have an extremely high Correlation. Hence, when either of these scores are entered on the “Predict” page, we use a Linear Regression model trained on the dataset to predict the other subject’s score.
-
Logistic Regression is used for the category of “Sports”. We find that students doing well in Math, are interested in sports as well and hence use a logistic regression model to predict whether a given student is interested in sports, using his performance in Math.
Uniqueness of the Application
The application lets the Teacher not only view the Student’s performance, but also lets the teacher estimate the performance of a class or a student in one subject using scores from another subject. Hence, the teacher can better understand and visualize the correlation between subjects. It also considers extracurriculars, such as “Interest in Sports”, as a factor that affects performance in various subjects.