2023-08-14 –, Aula
This tutorial offers a thorough introduction to the srai library for Geospatial Artificial Intelligence. Participants will learn how to use this library for geospatial tasks like downloading and processing OpenStreetMap data, extracting features from GTFS data, dividing an area into smaller regions, and representing regions in a vector space using various spatial features. Additionally, participants will learn to pre-train embedding models and train predictive models for downstream tasks.
In this tutorial, we intend to provide a comprehensive introduction to the Spatial Representations for Artificial Intelligence (srai) library. Participants will learn how to utilize this library for various geospatial applications, such as downloading and processing OpenStreetMap data, extracting features from GTFS data, splitting a given area into smaller regions, and embedding regions into a vector space based on different spatial features. Moreover, users will learn how to pre-train a model of their choice and build predictive models for use in downstream tasks.
By the end of the tutorial, attendees will be able to:
1. Install and set up the SRAI library.
2. Use SRAI to download and process geospatial data.
3. Apply various regionalization and embedding techniques to geospatial data.
4. Utilize pre-trained embedding models for clustering and similarity search.
5. Build predictive models on top of SRAI embeddings
6. Pre-train available models from scratch.
7. Understand the potential applications and future enhancements of the SRAI library.
If you want to follow along, please find the material and installation instructions at https://github.com/kraina-ai/srai-tutorial. We encourage you to set up the repository and install the dependencies before the tutorial.
Lastly, if you're not familiar with geospatial data, we can recommend a great tutorial by Joris Van den Bossche - Introduction to geospatial data analysis with GeoPandas.
It is not required to understand this tutorial, but it might allow you to build a deeper understanding of geospatial data and tooling in this domain. Consider it an optional pre-reading.
some
Public link to supporting material: Project Homepage / Git: Abstract as a tweet:Join us for an in-depth tutorial on the srai library, a powerful tool for GeoAI. Learn to handle geospatial data, create regions representations, and train specialized models
Category [Scientific Applications]:Geo Science
Expected audience expertise: Domain:some
Co-creator of SRAI library,
Passionate AI Researcher and ML Engineer working in NLP and GeoAI.
Graduated from the Wrocław University of Science and Technology with a Bachelor's in Computer Science and a Master's Degree in Data Science
Piotr Szymański is a scientist with a mathematical and computer science background. He obtained his Ph.D. in Computer Science at Wrocław University of Science and Technology in 2020. As a scholar, he visited Stanford University, Hasso Plattner Institute in Potsdam, Technical University of Sydney, Dortmund Technical University, and Josef Stefan Institute in Ljubljana. He is the primary author of the scikit-multilearn library for multi-label classification. He also has extensive corporate R&D experience. He was one of the authors of the ML/AI layers of Avaya Conversational Intelligence, a contact-center personnel support solution used widely in American call centers. Currently, he leads a Spatial AI group at the Department of Artificial Intelligence at Wrocław University of Science and Technology, Poland.
Co-creator of SRAI library. Master of Science in Data Science @ Wrocław University of Science and Technology. Machine Learning Engineer @ Brand24
Co-creator of SRAI library.
Spatial Data Scientist working at Allegro during the day and passionate open-source developer and geospatial researcher at night.
Graduated Master of Science in Data Science @ Wrocław University of Science and Technology.
Co-creator of the SRAI library,
An ML Engineer passionate about the geospatial domain and an author of highway2vec.
Background in Computer and Data Science from the Wrocław University of Science and Technology and a proud member of the KRAINA Lab tackling geospatial problems. MLOps Engineer @ GetInData