https://github.com/zazi2002/machine-learning-project

Introduction to Machine Learning Project
https://github.com/zazi2002/machine-learning-project

Last synced: 7 days ago
JSON representation

Introduction to Machine Learning Project

• Host: GitHub
• URL: https://github.com/zazi2002/machine-learning-project
• Owner: ZaZi2002
• Created: 2024-09-06T15:33:31.000Z (2 months ago)
• Default Branch: main
• Last Pushed: 2024-09-06T15:57:23.000Z (2 months ago)
• Last Synced: 2024-09-06T18:57:24.290Z (2 months ago)
• Language: Jupyter Notebook
• Homepage:
• Size: 404 KB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# Machine Learning Project: Dimensionality Reduction and Ensemble Learning

## Project Overview

This project is for Introduction to Machine Learning course and applies **dimensionality reduction** techniques, specifically **Principal Component Analysis (PCA)**, and uses **ensemble learning** methods such as **Random Forests** and **Decision Trees**. The goal is to improve the classification performance on a dataset by optimizing the number of features and weak learners. Key metrics such as accuracy, precision, recall, F1-score, and AUPRC (Area Under Precision-Recall Curve) are used to evaluate the model's performance.

### Key Features:

- **Dimensionality Reduction with PCA:** Reducing the number of features by maintaining the most variance-rich components.

- **Ensemble Learning:** Using multiple weak learners (Decision Trees) with both hard and soft voting strategies to enhance prediction accuracy.

- **Performance Metrics:** The model's output is evaluated using accuracy, precision, recall, F1-score, and AUPRC.

### Steps in the Notebook:

1. **Data Preprocessing:** 

   - Mean normalization and zero-centering of the data.

   - PCA to reduce the dimensionality of the dataset based on explained variance.

   

2. **Model Training:**

   - Train and test the model using the **Random Forest** estimator.

   - Implement ensemble learning with different numbers of weak learners.

   

3. **Performance Evaluation:**

   - Calculate key metrics including accuracy, precision, recall, F1-score, and AUPRC for both PCA-reduced data and ensemble learners.

   

4. **Optimization:** 

   - The number of PCA components and weak learners is optimized to balance performance and computational cost.

### Metrics Achieved:

- **Accuracy:** 97.7%

- **Precision:** 98.4%

- **Recall:** 98.7%

- **F1-Score:** 98.6%

- **AUPRC:** 98.1%

## Requirements

To run the project, you need the following Python libraries:

- `numpy`

- `scikit-learn`

- `matplotlib`