- Data Science Workspace overview
- Data Science Workspace access and features
- Data Science Workspace walkthrough
- JupyterLab notebooks
- Sensei Machine Learning API
- Pre-built recipes
- Model Authoring
- Model and recipe tutorials
- Create and publish a machine-learning model
- Create the demo Luma schemas and datasets
- Create the Retail Sales schema and dataset
- Preview schemas and datasets
- Package source files into a recipe
- Import a packaged recipe (UI)
- Import a packaged recipe (API)
- Train and evaluate a model (UI)
- Train and evaluate a model (API)
- Score a model (UI)
- Score a model (API)
- Optimize a model
- Enrich Real-Time Customer Profile with machine learning insights
- Publish a model as a service (UI)
- Publish a model as a service (API)
- Schedule automated training and scoring on a service (UI)
- Troubleshooting guide
- API reference
- Platform release notes
Data Science Workspace course
Data Science Workspace is no longer available for purchase.
This documentation is intended for existing customers with prior entitlements to Data Science Workspace.
This document provides a description of the expected learning outcomes in the Adobe Experience Platform Data Science Workspace course. In order to view the course, you must login to Experience League using your Adobe ID.
The getting started with Data Science Workspace for Data Scientists course is designed for data scientists who want to learn how to use JupyterLab Notebooks to derive insights and query data, create profile-enabled datasets, publish automated machine learning models, and activate machine-learned insights to both Adobe and non-Adobe applications.
Course prerequisites
- A registered Adobe ID account.
- The Adobe ID account must have been added to an Organization with access to Adobe Experience Platform and Data Science Workspace.
- A non-production sandbox.
Expected learning outcomes
The following learning outcomes are covered in the Data Science Workspace course. Additionally, you have the option to follow along while creating and publishing a propensity model that is provided for the course.
- The architecture of Data Science Workspace
- How to use JupyterLab
- How to access data and query data in Data Science Workspace
- Exploratory Data Analysis
- How to create a recipe and model
- Methods used to train and score a model
- The role of hyper-parameters in model development
- How to publish trained models as a service
- How to use Data Science Workspace to enrich your Real-Time Customer Profile data
- How to create a streaming segment with your model output
Lessons
The Data Science Workspace course is split into five lessons.
Lesson 1
Introduction (19 minutes): Learn about the course and get a high-level overview of Data Science Workspace including the required course assets.
Lesson 2
Load, query, and explore data in JupyterLab (24 minutes): Learn how JupyterLab on Experience Platform helps simplify and facilitate key workflows for a data scientist, such as gathering data, cleaning data, visualizing data, and discovering insights.
Lesson 3
Create a model in JupyterLab (26 minutes): Learn how to start building models in Data Science Workspace.
Lesson 4
Use Data Science Workspace to train and score a model (6 minutes): Learn how to create a model and publish it as a service in Experience Platform.
Lesson 5
Consume and deliver Data Science Insights (11 minutes): Learn how Data Science Workspace model outputs can be used in the Real-Time Customer Profile to deliver personalized experiences with Adobe applications and services.
Next steps
After completing the Data Science Workspace course, visit the Sensei Machine Learning API guides to learn how to utilize RESTful APIs to do everything you just learned and more.
