GPA Predictor and Academic Advisor Web App
Project Overview This project aims to create a web application that enables individuals in the academic community, including students, teachers, and parents, to predict grade point averages (GPA) based on essential parameters such as:
- Study Hours
- Previous GPA
- Physical Activity
- Stress Levels
- Educational Resources
- Nutrition
- Sleep Patterns
- Time Wasted on Social Media
Methodology
Our approach involves training an artificial neural network (ANN) with relevant data to predict GPA values. These predicted values are then passed to a pre-trained Open-source "Cohere" language model(Command-Xlarge API) to provide personalized advice based on the relevant features influencing the GPA prediction. The advice is displayed in the following format:
General Impression "Based on the information provided, it seems..."
Academic Performance Study Hours: While studying for more hours is beneficial, ensure you maintain a balance to avoid burnout. Use of Available Resources: Despite limited educational resources, maximizing their use can significantly impact your performance.
Well-being Improve Nutrition: A low level of nutrition can negatively impact your academic performance. Aim for a balanced diet. Increase Physical Activity: Aim to increase your physical activity to boost both physical and mental health. Limit Social Media Time: Excessive time on social media can detract from study time and impact your grades.
Goal Setting ... Hosting the Application on Streamlit Cloud
- Sign in to Streamlit.io.
- Link your GitHub account with Streamlit after signing up.
- Click on 'Create New Application' at the top left of the screen.
- Select the public repository (link provided below).
- Choose the
finalapp.pyfile in the section where the application file is required. - Edit the URL as needed, then click on the deploy button to deploy the application. Conclusion Executing this project involved training various models, with the ANN proving to be the best. All required files have been uploaded to the repository. Below is a link to our YouTube video demonstrating how our platform works: https://youtu.be/IilmEGuVado