Partitioned Models (Open)

Learn how to train a machine learning model on partitions of your dataset. (Open versions of courses do not count towards the certificate program.)

Partitioned Models (Open)


About this course

About this course

To track your progress and work towards completion of the certificate program, you should register for this course through the Learning Paths.
Any progress in this open version of the course will not count towards a certificate.


Partitioned Models demonstrates how to build a partitioned model trained on subgroups, or partitions, of the dataset, and compare the results with a non-partitioned model trained on the whole dataset.


In the hands-on lesson, we will work with flight data. We will predict the arrival delay time for a given flight based on the flight's characteristics, such as departure delay time. Before the hands-on section, you will have a chance to grasp the concept of partitioned models by watching short videos.




Learning Objectives

At the end of this course, you will be able to:

1 - Describe partitioned models to a colleague

2 - Create a partitioned model from a partitioned dataset in Dataiku DSS

3 - Compare results from a non-partitioned model with the partitioned model


Course Properties

Course Title Partitioned Models

Target Audience

Anyone who wants to learn how to create partitioned models in Dataiku DSS

Access Level

Free / included with registration

Estimated Time for Completion

24 Minutes

Completion Criteria

Pass the course checkpoint with 80%

Supplemental Materials (Y/N)


Knowledge Prerequisite(s)


Technical Prerequisite(s)

Dataiku DSS - Latest version (FREE EDITION is enough)


Start by watching the Partitioned Models course overview video.