universal-ml-framework 1.0.0

A complete machine learning pipeline framework for any dataset

🌟 Universal ML Framework

A complete, automated machine learning pipeline framework that works with any dataset. Build, compare, and deploy ML models with minimal code.

🚀 Key Features

• 🤖 Automated Pipeline - Complete ML workflow from data to deployment
• 🔍 Auto Feature Detection - Automatically detects numeric, categorical, and binary features
• 📊 Model Comparison - Compares multiple algorithms with cross-validation
• ⚙️ Hyperparameter Tuning - Automatic parameter optimization
• 🎯 Multi-Problem Support - Classification and regression tasks
• 📦 Production Ready - Model persistence and metadata tracking

📦 Installation

# Install from source
git clone https://github.com/yourusername/universal-ml-framework.git
cd universal-ml-framework
pip install -e .

# Or install dependencies manually
pip install pandas scikit-learn numpy joblib

🎯 Quick Start

Basic Usage

from universal_ml_framework import UniversalMLPipeline

# Classification
pipeline = UniversalMLPipeline(problem_type='classification')
pipeline.run_pipeline(
    train_path='data.csv',
    target_column='target',
    test_path='test.csv'
)

# Regression
pipeline = UniversalMLPipeline(problem_type='regression')
pipeline.run_pipeline(
    train_path='data.csv',
    target_column='price'
)

Quick Setup Functions

from universal_ml_framework import quick_classification_pipeline

# One-liner for classification
result = quick_classification_pipeline('data.csv', 'target_column')

Generate Sample Data

from universal_ml_framework import DataGenerator

# Generate synthetic datasets for testing
DataGenerator.generate_customer_churn()
DataGenerator.generate_house_prices()
DataGenerator.generate_sales_forecasting()

🔧 Supported Algorithms

Classification

• Random Forest Classifier
• Logistic Regression
• Support Vector Machine

Regression

• Random Forest Regressor
• Linear Regression
• Support Vector Regression

📊 What It Does

1. Data Loading - Reads CSV files automatically
2. Feature Detection - Identifies feature types (numeric/categorical/binary)
3. Preprocessing - Handles missing values, encoding, scaling
4. Model Training - Trains multiple algorithms with cross-validation
5. Hyperparameter Tuning - Optimizes best performing model
6. Prediction - Generates predictions on test data
7. Model Saving - Persists trained model and metadata

📈 Output Files

• predictions.csv - Test set predictions
• best_model.pkl - Trained model (joblib format)
• model_info.json - Model metadata and performance

🎛️ Customization

Custom Feature Types

pipeline.feature_types = {
    'numeric': ['age', 'income'],
    'categorical': ['city', 'category'],
    'binary': ['has_feature']
}

Exclude Columns

pipeline.run_pipeline(
    train_path='data.csv',
    target_column='target',
    exclude_columns=['id', 'timestamp']
)

🧪 Demo

# Run complete demo with synthetic data
python demo.py

📁 Project Structure

universal_ml_framework/
├── core/
│   └── pipeline.py          # Main pipeline class
├── utils/
│   ├── helpers.py           # Helper functions
│   └── data_generator.py    # Synthetic data generation
├── configs/
│   └── dataset_configs.py   # Predefined configurations
└── examples/
    ├── basic_usage.py       # Basic examples
    └── advanced_usage.py    # Advanced examples

🎯 Use Cases

• Business Analytics - Customer churn, sales forecasting
• Finance - Credit risk, fraud detection
• Healthcare - Medical diagnosis, treatment prediction
• Marketing - Campaign response, customer segmentation
• Real Estate - Price prediction, market analysis
• HR - Employee performance, retention prediction

🤝 Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

🙏 Acknowledgments

• Built with scikit-learn
• Powered by pandas and numpy