Version 0.2.5

We add the following new features to this version:

• Full pipeline GPU acceleration

  • Data adaption

  • Data cleaning

  • Feature selection

  • Data drift detection

  • Feature selection(2nd stage)

  • Pseudo labeling(2nd stage)

  • Optimization

    • Data preprocessing

    • Model fitting

  • Model ensemble

  • Metrics

• Model training

  • Add TargetEncoder for categories

  • Set estimator eval_metric based on experiment reward_metric

• Advanced Features

  • Data adaption in experiment

• Experiment Visualization

  • Experiment configurations

  • Dataset information

  • Processing information

• Multijob management

  • Series and parallel jobs scheduling

  • Local and remote jobs execution

• Export experiment report