Spatial Data Science for Business


Explore the application of spatial data science to uncover hidden patterns and improve predictive modeling.

Thierry Warin (HEC Montréal and CIRANO (Canada))

Course overview

Spatial data science allows analysts to extract deeper insight from data using a comprehensive set of analytical methods and spatial algorithms, including machine learning and deep learning techniques. This course explores the application of spatial data science to uncover hidden patterns and improve predictive modeling. You’ll work with powerful analytical tools in Esri’s ArcGIS software and learn how to integrate popular open data science packages into your analyses.

Topics covered

Educational Objectives




[Unfold some info by clicking on the triangle pictogram]

Session 1: Introduction to Spatial Data Science here

Explore how spatial data, tools, and analysis techniques augment traditional data science. Understand that “spatial” means more than x,y coordinates and that place-based context reveals patterns in data that otherwise may be hidden. Start applying data engineering and visualization techniques in ArcGIS Pro and ArcGIS Notebooks.

Session 2: The Spatial Approach to Predictive Analysis here

Prediction is fundamental to data science. See how incorporating spatial properties into mod- eling workflows deepens understanding of data and adds predictive power. Learn how to apply random forest, a widely used machine learning approach, to solve problems. Train and evaluate a model, then use it to generate robust predictions.

Session 3: Finding Optimal Locations Using Suitability Models here

Apply widely used spatial analysis techniques to answer this universal question asked by all kinds of organizations: Where is the best location for ? Perform a weighted overlay analysis that considers and ranks multiple suitability criteria. Learn how to transform data using functions to more completely represent suitability impact.

Session 4: Pattern Detection and Clustering here

Does a pattern have meaning or is it a product of random chance? ArcGIS includes a suite of tools to help analysts identify patterns and clusters in data and determine if they are meaningful. Learn how to apply statistical clustering methods to analyze patterns in space as well as time. Create a space- time cube, then use space-time pattern mining tools to explore spatiotemporal trends and determine where and when high and low clusters occur.

Session 5: Object Detection with Deep Learning here

Take a deep dive into extracting information from massive data using deep learning. Learn how to automate the process of detecting objects and identifying features from imagery. Practice pre- paring training sample data, then use a neural network to train an object detection model.

Session 6: Communicating Results with Impact here

Successful analysts and data scientists must be effective storytellers. Learn how to present a complex analysis using information products that resonate with your audience. Design and build inter- active, visual stories that share the key information you want decision-makers to understand and act on.


Before class:

  1. Read this syllabus thoroughly
  2. Bookmark this page:

That’s it!


For attribution, please cite this work as

Warin (2020, Jan. 1). Spatial Data Science for Business. Retrieved from

BibTeX citation

  author = {Warin, Thierry},
  title = { Spatial Data Science for Business},
  url = {},
  year = {2020}