Student Academic Performance Prediction System Using AI&ML

Abstract

Predicting student performance using AI and 
ML techniques has gained significant attention as a 
means to enhance educational outcomes and provide 
better support for learners. This report explores the 
development and implementation of a predictive system 
that estimates student performance based on user
inputted factors such as sleep hours and study duration. 
The system leverages a linear regression model to 
analyze key attributes like study habits and academic 
history, offering accurate predictions of student 
outcomes. Unlike traditional methods that rely on 
predefined datasets, this system processes real-time 
user-provided data. Additionally, Firebase is utilized for 
efficient data storage and management, while the 
Google Gemini API enhances predictive accuracy and 
user interaction. The system’s effectiveness has been 
validated 
through 
experimental 
evaluations, 
demonstrating its capability to predict academic 
performance metrics, such as GPA or exam scores, with 
high reliability. This study contributes to the field of 
educational analytics by providing insights that can be 
used to support personalized learning strategies and 
informed decision-making in academic settings. By 
continuously refining and expanding its features, the 
system has the potential to improve educational 
practices and foster student success in diverse learning 
environments.