Supervised Learning Project

This repository contains code for a supervised learning project aimed at predicting student GPA using various machine learning algorithms.

NOTE: The error in the ipynb file is there to show that the dataset must be downloaded before running the file. Once downloaded everything will behave as intended.

Project Overview

• Objective: Predict student GPA based on multiple features using different regression models.
• Algorithms Used:
  • Linear Regression: Used to establish a baseline performance by fitting a linear equation to the observed data.
  • Random Forest Regression: Applied to capture non-linear relationships, providing an ensemble learning method to improve prediction accuracy.
  • K-Nearest Neighbors (KNN) Regression: Employed to predict values based on the average of the 'k' closest data points in the feature space.

Methodology

• Data Preparation: Selected features from dataset, excluding 'id' and 'gpa' for training.
• Model Training and Testing: Each model was trained on 80% of the data with the remaining 20% reserved for testing.
• Model Evaluation: Performance assessed using Mean Squared Error (MSE).

Usage

To run this project:

• Download the dataset listed at the top of the .ipynb file
• Make sure it is in a reachable location

git clone https://github.com/LIoccoUMD/supervised-learning/
cd supervised-learning
python Supervised_Learning.ipynb

OR, for an easier setup

• Open in Google Colab
• Download the linked dataset
• Run directly in Colab