This event is part of the Harvard Affiliate Only Spatial Data Science Workshop Series.
Date: 1/28/2022Time: 12:00 PM ETLength: 1 hour of presentation with an hour for open discussion/office hours afterwardsRegistration link.
Machine learning can play a critical role in spatial problem solving in a wide range of areas, including spatial pattern detection. The ability to automate this type of pattern detection is crucial when addressing the world’s most challenging spatial problems. This workshop covers machine learning methods to perform cluster analysis using attributes, space, and time. We'll cover methods such as Multivariate Clustering, Build Balanced Zones, and Density-based Clustering in both space and time. Each technique is explained in conceptual terms to understand how it works and accompanied by a demonstration that showcases its use. Use cases will include climate and environmental data, data for social good, and more.
Prior to the workshop attendees will be provided an ArcGIS Pro project with demo data that will be used throughout the workshop. Attendees will be invited to follow along during demonstrations using the provided data, though this is not required.Learning Objectives:
To introduce the various challenges that create a need for data-driven clustering and categorization of data
To see how machine learning concepts are integrated into spatial analysis workflows in GIS
To learn the concepts behind common density-based clustering algorithms, including DBSCAN, HDBSCAN, and OPTICS
To learn the concepts behind attribute clustering algorithms, including k-means and Spatially Constrained Multivariate Clustering (SKATER)
To understand the implications of the various parameters of these methods
To be able to visualize and interpret the results of cluster analysis in maps
Basic experience in ArcGIS Pro’s functionality, including the ability to create and open projects, load layers, and use geoprocessing tools.
Anyone that wants to visualize and analyze continuous data spatially
Anyone that wants to use maps and spatial data in their work