automac 0.1.2

An all-in-one automated ML pipeline for advanced feature engineering, Boruta selection, and parallel optimization.

📦 Automac: High-Performance Automated ML Pipeline

Automac ek end-to-end Automated Machine Learning (AutoML) library hai jo data preprocessing se lekar model optimization tak ka saara heavy lifting khud karti hai. Isme advanced techniques jaise Boruta Feature Selection aur Optuna-based Parallel Tuning integrated hain.

---

🚀 Key Modules & Features

1. 🛡️ Advanced Feature Engineering

Sirf scaling nahi, balki statistically solid feature selection.

• Boruta Selection: Shadow features ke saath compete karke irrelevant noise ko hatana.
• Smart Handling: Automatic outlier clipping (IQR) aur multicollinearity removal.
• Encoding: High-cardinality data ke liye advanced Target Encoding.

2. ⚡ Automated Model Tuning

Parallel execution jo aapke CPU ke har core ka sahi istemaal karti hai.

• Optuna Integration: Hyperparameter optimization ka gold standard.
• Smart Allocation: Cores ko models ke beech distribute karna taaki Windows/Linux dono par maximum speed mile.
• Supported Models: XGBoost, LightGBM, CatBoost, RandomForest, SVM, KNN, etc.

3. 📝 Text Preprocessing (NLP)

Raw text data ko cleaning aur normalization ke liye ready karna.

• Stopword removal, regex-based tokenization, aur Porter Stemming.

4. 📊 Diagnostics & Visualization

Model ko "Black Box" banne se rokna.

• Learning Curves: Training vs Validation lines se Overfitting detect karna.

---

🛠️ Installation

# Clone the repository
git clone [https://github.com/jubito-27/ml-automator.git](https://github.com/jubito-27/ml-automator.git)
cd ml-automator

# Install in editable mode
pip install -e .