ml-mini-project

Mini-project for UE23CS352A Machine Learning – Random Forest Model for Predicting Coral Reef Regimes from Human and Natural Influences

Problem Statement

This project aims to predict coral reef ecosystem regimes using data on human activities and natural environmental factors. The goal is to classify reefs into different regime types based on these influencing variables.

Approach

1. Data Preprocessing – Cleaned and normalized the input dataset.
2. Feature Selection – Selected key features affecting coral reef regimes.
3. Model Training – Trained multiple models (e.g., Random Forest ).
4. Evaluation – Compared models using F1-score, accuracy, and confusion matrix.
5. Prediction – Saved the final model and generated predictions on test data.

Results

• Final Model: Random Forest Classifier
• F1-Score: 0.9677(Micro-averageed F1 Score)
• Accuracy: 96.77%

Installation & Usage

Clone this repository: cd "C:\Users\archi\OneDrive\Desktop\5thsem\ML\ml_mini_project\ml-mini-project" dir pip install -r requirements.txt pip install notebook jupyter notebook

Team Members

• Apoorva Biradar (PES1UG23CS095)
• Archi Pankaj (PES1UG23CS099)