AutoMLBench: Automating Machine Learning Workflows

Automated Machine Learning Benchmarking Library

🚀 AutoMLBench provides a seamless way to compare machine learning models, preprocess data, evaluate performance, and optimize models with hyperparameter tuning.

---

📌 Installation

Ensure all dependencies are installed:

pip install pandas scikit-learn numpy matplotlib xgboost lightgbm catboost imbalanced-learn

Install from pypi:

pip install automlbench

For local development:

git clone https://github.com/AnnNaserNabil/automlbench.git
cd automlbench
pip install -e .

Model Comparison Without Hyperparameter Tuning

The simplest way to compare models using AutoMLBench.

1️⃣ Load Dataset & Preprocess

import pandas as pd
from automlbench import preprocess_data, get_models, train_models, evaluate_model, plot_performance

# Load dataset
url = "DATAPATH"
df = pd.read_csv(url) 

# Define target column
target_column = "Name OF the Target Column"

# Preprocess data
X_train, X_test, y_train, y_test = preprocess_data(df, target_column)

---

2️⃣ Train All Default Models

# Get predefined models
models = get_models()

# Train models without tuning
results = train_models(X_train, X_test, y_train, y_test)

# Print model performance results
print(results)

---

3️⃣ Evaluate & Compare Model Performance

# Evaluate all models
for model_name, model in models.items():
    print(f"Evaluating {model_name}...")
    metrics = evaluate_model(model.fit(X_train, y_train), X_test, y_test)
    print(metrics)

# Plot performance comparison
plot_performance(results, metrics=["Accuracy", "Precision", "Recall", "F1-Score", "RMSE"])

---

🔹 Model Comparison With Hyperparameter Tuning

For better performance, use hyperparameter tuning.

1️⃣ Get Hyperparameter Grids

from automlbench import get_hyperparameter_grids, tune_hyperparameters

# Retrieve hyperparameter grids
hyperparameter_grids = get_hyperparameter_grids()

---

2️⃣ Tune Models

best_models = {}

# Tune each model if it has a predefined hyperparameter grid
for model_name, model in models.items():
    if model_name in hyperparameter_grids:
        print(f"Tuning {model_name}...")
        best_model, best_params = tune_hyperparameters(model, hyperparameter_grids[model_name], X_train, y_train)
        best_models[model_name] = best_model
        print(f"Best params for {model_name}: {best_params}")
    else:
        best_models[model_name] = model  # Use default if no tuning grid

---

3️⃣ Train Tuned Models

# Train models using the best hyperparameters found
tuned_results = train_models(
    X_train, X_test, y_train, y_test, 
    selected_models=list(best_models.keys()), 
    hyperparams={name: model.get_params() for name, model in best_models.items()}
)

# Display tuned model results
print(tuned_results)

---

4️⃣ Evaluate & Compare Tuned Models

# Evaluate all tuned models
for model_name, model in best_models.items():
    print(f"Evaluating {model_name}...")
    metrics = evaluate_model(model.fit(X_train, y_train), X_test, y_test)
    print(metrics)

# Plot comparison of tuned models
plot_performance(tuned_results, metrics=["Accuracy", "Precision", "Recall", "F1-Score", "RMSE"])

⚡ Quick Summary

✅ Basic Comparison – Train models with default settings.
✅ Hyperparameter Tuning – Optimize models for better performance.
✅ Evaluation & Visualization – Compare accuracy, precision, recall, F1-score, and RMSE.
✅ Automated ML Benchmarking – Quickly assess multiple models with minimal code.

---

📌 Contributing

Contributions are welcome! To contribute:

1. Fork the repository
2. Create a new branch (feature-branch)
3. Make changes & test (pytest tests/)
4. Submit a pull request (PR)

---

📜 License

AutoMLBench is released under the MIT License.

---

Get Started Today!

Try AutoMLBench and streamline your machine learning projects. Visit the GitHub repository for more details, examples, and contributions. 🚀 ```

This version includes the appropriate metadata and a refined introduction to AutoMLBench. Let me know if you’d like any modifications! 🚀