Visual Machine Learning and Modeling in Dataiku

READ DATASHEET
  • Step-by-step guided
    machine learning
  • Scalable and customizable
    machine learning
  • Fully production
    ready
  • Step-by-step guided
    machine learning
  • Scalable and customizable
    machine learning
  • Fully production
    ready
Product screenshot showing machine learning models and their respective accuracies on a visual UI interface

Step-by-step guided machine learning

  • Clean data, create new features, and build your model in a unified environment.
  • Get instant visual feedback to assess your model’s performance.
  • Compare and optimize your models using various cross validation strategies.
Product screenshot showing the features having the most impact on a machine learning model

Understand your models

  • Immediately see which features have the most impact on your prediction with variable importance.
  • Quickly understand complex feature interactions and analyse coefficients.
  • Automated visual and statistical reports help you interpret the clusters resulting from non-supervised machine learning.
Product screenshot showing the visual UI to leverage machine learning models

Use visual UI or code to leverage latest ML technologies

  • The visual machine learning in DSS leverages state-of-the-art machine learning libraries: Scikit-Learn, MLlib, XGboost.
  • Customize code directly in Python and R for advanced custom machine learning.
  • Integrate any external machine learning library accessible through code APIs (H2O, Dato, Skytree, etc).
Product screenshot showing the visual UI to assess machine learning models

Fully production ready

As soon as your model is built and assessed, instantly:

  • Use it for batch scoring within your data workflow.
  • Deploy it as a real-time prediction service (REST API).
  • Manage your model lifetime: save previous versions of a retrained model and rollback to any secure version in just one click.
  • Manage your model's performance with a feedback loop.