What is AutoML

Onepanel's AutoML framework was built to improve the accuracy of your machine learning models and make them more accessible by automatically creating a data analysis pipeline that can include data pre-processing, feature selection, and feature engineering methods along with machine learning methods and parameter settings that are optimized for your data. Each of these steps are usually very time-consuming for the machine learning author. 

Onepanel's AutoML is installed automatically when you start a job or a notebook.  You can access the library by importing the 'automl.' library similar to the example below:

from automl.pipeline import LocalExecutor, Pipeline, PipelineStep, PipelineData
from automl.data.dataset import Dataset
from automl.model import ModelSpace, CV, Validate, ChooseBest
from automl.hyperparam.templates import (random_forest_hp_space,
                                         knn_hp_space, svc_kernel_hp_space,
from automl.feature.generators import FormulaFeatureGenerator, PolynomialGenerator
from automl.feature.selector import FeatureSelector
from automl.hyperparam.optimization import Hyperopt
from automl.combinators import RandomChoice


  • Flexible UI for solving typical machine learning tasks
  • Python framework with high-level DSL (Domain Specific Language) for more complex pipelines
  • Easily extensible for different machine learning models and data formats
  • Run pipelines on a single machine or in a cluster (coming soon)


  • Extensibility
  • Ease of use 
  • Pipeline algorithm and execution strategy are separate by design
  • Scikit-Learn integration 
  • XGBoost integration 
  • Hyperopt integration

AutoML Pipeline

AutoML Example:

Onepanel has applied our AutoML framework on the Kaggle Titantic competition ( https://www.kaggle.com/c/titanic ) documented in an easy to walkthrough Jupyter Notebook.

Onepanel Kaggle Titantic Competition Example:


Did this answer your question?