Machine Learning Projects Repository 🚀

Welcome to my comprehensive collection of machine learning and computer vision projects! This repository showcases a diverse range of projects from data science and machine learning to computer vision applications using YOLO and deep learning frameworks.

📋 Table of Contents

• Overview
• Projects
  • Machine Learning Projects
  • Computer Vision Projects
  • Competition Projects
• Technologies Used
• Setup Instructions
• Project Structure
• Contributing

🔍 Overview

This repository contains 9 different projects covering various aspects of machine learning, deep learning, and computer vision. Each project demonstrates different techniques, from traditional machine learning algorithms to state-of-the-art deep learning models and real-time object detection systems.

🎯 Projects

Machine Learning Projects

1. 🚜 Bulldozer Price Prediction

Location: Bulldozer-price-prediction/

A comprehensive end-to-end machine learning project that predicts the sale price of bulldozers using regression techniques.

• Problem: Predict future sale prices of bulldozers based on their characteristics and historical data
• Dataset: Kaggle Bluebook for Bulldozers competition data
• Evaluation Metric: RMSLE (Root Mean Squared Log Error)
• Features: Multiple bulldozer characteristics and specifications
• Key Files:
  • end-to-end-Bulldozer-price-regression.ipynb - Main analysis notebook
  • Final_predictions.csv - Model predictions
  • environment.yml - Conda environment configuration
  • dataset/ - Training, validation, and test datasets

2. ❤️ Heart Disease Classification

Location: Heart-disease-project/

A binary classification project to predict heart disease presence based on clinical parameters.

• Problem: Predict whether a patient has heart disease based on medical attributes
• Dataset: Cleveland data from UCI Machine Learning Repository
• Target Accuracy: 95% for proof of concept
• Features: Age, sex, chest pain type, blood pressure, cholesterol, ECG results, etc.
• Key Files:
  • end-to-end-heart-disease-classification.ipynb - Complete analysis
  • GridSearchCV_LogisticRegresion.pkl - Trained model
  • environment.yml - Dependencies
  • data/heart-disease.csv - Dataset

3. 🐕 Dog Breed Prediction

Location: Dog-breed-prediction/

A multi-class image classification project using deep learning to identify dog breeds.

• Problem: Classify dog breeds from images using computer vision
• Model: MobileNetV2 with transfer learning
• Framework: TensorFlow/Keras
• Key Files:
  • end_to_end_dog_vision.ipynb - Main notebook
  • models/ - Trained model files (.h5 format)
  • logs/ - Training logs and TensorBoard data
  • custom-data/ - Custom test images
  • full_model_predictions_submission_1_mobilenetV2.csv - Predictions

4. 🖐️ Palmprint Detection

Location: Palmprint-detection-project/

A specialized computer vision project for palmprint detection and analysis.

• Research Paper: Included (9939-Article Text-17682-1-10-20210714.pdf)
• Key Files:
  • end_to_end_palmprint_detection.ipynb - Implementation notebook
  • custom-images/ - Test palmprint images
  • environment.yml - Environment setup

Computer Vision Projects

5. 🚗 Car Counter (Project 1)

Location: Project 1 - Car Counter/

Real-time vehicle counting system using YOLO object detection and SORT tracking.

• Technology: YOLOv8, OpenCV, SORT tracking algorithm
• Features: Real-time detection, vehicle tracking, counting across a line
• Key Files:
  • Car-Counter.py - Main application
  • sort.py - SORT tracking implementation
  • mask.png - Region of interest mask
  • graphics.png - UI graphics

6. 👥 People Counter (Project 2)

Location: Project 2 - People Counter/

Similar to the car counter but optimized for counting people in video streams.

• Technology: YOLO, OpenCV, tracking algorithms
• Application: Crowd monitoring, occupancy counting
• Key Files:
  • People-Counter.py - Main application
  • sort.py - Tracking algorithm
  • mask.png, graphics.png - UI components

7. 🦺 PPE Detection (Project 3)

Location: Project 3 - PPE Detection/

Personal Protective Equipment detection system for workplace safety monitoring.

• Purpose: Detect safety equipment like hardhats, masks, safety vests
• Classes: 25 different classes including safety equipment and vehicles
• Key Files:
  • PPEDetection.py - Main detection script
  • best.pt - Trained YOLO model weights

Detected Classes:

• Safety Equipment: Hardhat, Mask, Safety Vest, Gloves
• Violations: NO-Hardhat, NO-Mask, NO-Safety Vest
• Vehicles: Excavator, SUV, truck, bus, etc.
• Other: Person, Safety Cone, Ladder, etc.

8. 🃏 Poker Card Detection (Project 4)

Location: Project 4 - Poker Card Detection/

Advanced computer vision system for detecting and recognizing poker cards and hands.

• Features: Card detection, hand recognition, poker hand classification
• Key Files:
  • Poker-Hand-Detector.py - Main detector
  • PokerHandFunction.py - Hand evaluation logic
  • best.pt, last.pt - Model weights
  • Training_results/ - Training metrics and validation images

Competition Projects

9. 🌍 Smartathon HackerEarth - Pollutant Classification

Location: Smartathon-Hackerearth/

A machine learning competition project for environmental data analysis.

• Competition: HackerEarth Smartathon
• Problem: Pollutant classification from environmental data
• Duration: 10-day learning and implementation project
• Key Files:
  • Pollutant_classification.ipynb - Complete solution
  • readme.md - Project-specific documentation
• Dataset: Available via Google Drive link in project readme

🛠️ Technologies Used

Machine Learning & Data Science

• Python Libraries: pandas, numpy, scikit-learn, matplotlib, seaborn
• Machine Learning: Random Forest, Logistic Regression, GridSearchCV
• Deep Learning: TensorFlow, Keras, MobileNetV2
• Environment Management: Conda, pip

Computer Vision

• Frameworks: OpenCV, Ultralytics YOLO (YOLOv8)
• Object Detection: YOLO models, custom trained weights
• Tracking: SORT (Simple Online and Realtime Tracking)
• Image Processing: cvzone, cv2

Development Tools

• Notebooks: Jupyter Notebook (.ipynb)
• Version Control: Git
• Model Formats: .h5 (Keras), .pt (PyTorch/YOLO), .pkl (scikit-learn)

🚀 Setup Instructions

General Requirements

# Clone the repository
git clone https://github.com/Rahul-lalwani-learner/Machine-learning-projects.git
cd Machine-learning-projects

For Machine Learning Projects

# Navigate to specific project directory
cd Heart-disease-project  # or Bulldozer-price-prediction, etc.

# Create conda environment
conda env create -f environment.yml
conda activate [environment-name]

# Launch Jupyter Notebook
jupyter notebook

For Computer Vision Projects

# Install required packages
pip install ultralytics opencv-python cvzone

# For YOLO projects, ensure you have the model weights
# Some projects include pre-trained weights (best.pt)
# Others may require downloading YOLOv8 weights

# Run the application
python Car-Counter.py  # or respective project file

Video Files and Weights

• Video files are referenced from ../Videos/ directory
• YOLO weights from ../Yolo-Weights/ for some projects
• Some projects include custom trained weights (best.pt, last.pt)

📁 Project Structure

Machine-learning-projects/
├── README.md (this file)
├── Bulldozer-price-prediction/
│   ├── end-to-end-Bulldozer-price-regression.ipynb
│   ├── environment.yml
│   ├── Final_predictions.csv
│   └── dataset/ (training data)
├── Dog-breed-prediction/
│   ├── end_to_end_dog_vision.ipynb
│   ├── models/ (trained models)
│   ├── logs/ (training logs)
│   └── custom-data/ (test images)
├── Heart-disease-project/
│   ├── end-to-end-heart-disease-classification.ipynb
│   ├── GridSearchCV_LogisticRegresion.pkl
│   └── data/heart-disease.csv
├── Palmprint-detection-project/
│   ├── end_to_end_palmprint_detection.ipynb
│   └── custom-images/
├── Project 1 - Car Counter/
│   ├── Car-Counter.py
│   └── sort.py
├── Project 2 - People Counter/
│   ├── People-Counter.py
│   └── sort.py
├── Project 3 - PPE Detection/
│   ├── PPEDetection.py
│   └── best.pt
├── Project 4 - Poker Card Detection/
│   ├── Poker-Hand-Detector.py
│   ├── PokerHandFunction.py
│   ├── best.pt
│   └── Training_results/
└── Smartathon-Hackerearth/
    ├── Pollutant_classification.ipynb
    └── readme.md

🎯 Key Features

Machine Learning Projects

• End-to-end workflows from data preprocessing to model deployment
• Multiple evaluation metrics and model comparison
• Feature engineering and data visualization
• Model persistence with pickle and joblib
• Environment reproducibility with conda/pip requirements

Computer Vision Projects

• Real-time processing capabilities
• Object tracking across video frames
• Custom trained models for specific use cases
• Multi-class detection with confidence scoring
• Visual feedback with bounding boxes and labels

📊 Results and Performance

Machine Learning

• Bulldozer Price Prediction: RMSLE optimization for auction price prediction
• Heart Disease Classification: Target 95% accuracy achieved
• Dog Breed Prediction: Multi-class classification with MobileNetV2
• Palmprint Detection: Specialized biometric recognition

Computer Vision

• Car/People Counting: Real-time tracking and counting
• PPE Detection: 25-class safety equipment detection
• Poker Card Detection: Complex card and hand recognition

🤝 Contributing

Feel free to contribute to any of these projects by:

1. Fork the repository
2. Create a feature branch
3. Make your improvements
4. Submit a pull request

📝 Notes

• Some projects require additional dataset downloads (links provided in respective project folders)
• Video files and some model weights may need to be downloaded separately
• Environment files are provided for reproducible setups
• Each project includes detailed notebooks with step-by-step explanations

📧 Contact

For questions, suggestions, or collaboration opportunities, please feel free to reach out!

---

Happy Learning and Coding! 🚀