2025-02-25, 15:45–17:15, Room 2
This tutorial will present a full end-to-end example for the site selection task (Where to place bicycle-sharing stations?) using the machine-learning model and features based on data from OpenStreetMap and Overture Maps.
We will focus on available methods for feature engineering with geospatial data and how to create a full pipeline from acquiring the data for the region of interest to the final prediction for the site selection task (where to place bicycle-sharing stations).
Most of the tutorial will utilize the srai library, which is dedicated to geospatial machine learning.
Attendees will learn how the OSM and OvertureMaps data is structured, how to download the data, use H3 grid to tesellate the area, prepare features, train the model with spatial stratification and evaluate the results (with nice map-based visualizations).
Tutorial materials: https://github.com/kraina-ai/srai-tutorial/tree/geopython2025