Skip to content

AutoDask

Note: AutoDask is currently in the development stage, which is why some modules may not work or may have errors.

The Distributed AutoML

AutoDask is a lightweight AutoML library that brings together the power of distributed computing with Dask and the intelligence of Bee Colony Optimization for hyperparameter tuning.

pip install autodask  <unsupported yet>

from autodask.main import AutoDask

# Create an AutoDask instance
adsk = AutoDask(task='classification')

# Train the model
adsk.fit(X_train, y_train)

# Make predictions
predictions = adsk.predict(X_test)

Core Features

  • Multi-Task Support: Classification and regression workflows
  • Distributed Computing: Parallel model training and evaluation
  • Automated Feature Engineering: Intelligent preprocessing and transformation
  • Hyperparameter Optimization: Nature-inspired Bee Colony Optimization algorithm
  • Model Ensembling: Combines top-performing models by using weighted average blending

Example Use Cases

Coming soon...