👉 The slide deck for this talk → https://docs.google.com/presentation/d/1jT2biH6DIlN133djneZcdXS7-CgR-dlWm2jBQ1KhM50/edit?usp=sharing
🛰️ More about the 2023 Geo for Good Summit → https://g.co/earth/geoforgood23
❤️🌍 DESCRIPTION:
Unlock the full potential of geospatial data with our deep dive into Earth Engine classifiers! 🌍🔍 In this enlightening session, Noel and Emma guide you through the intricacies of machine learning, revealing how to master Earth Engine's built-in classifiers for landcover mapping and more. Whether you're sampling points with Landsat data, navigating the complexities of Random Forest, or exploring ensembles, this video is your ultimate resource. Learn more about creating accurate maps, avoiding common pitfalls, and employing advanced techniques like stratified sampling and texture metrics. Ready to harness Earth's secrets for a sustainable future? Dive into our video now and reuse the code to make a difference! #GeospatialData #MachineLearning #EarthEngine #LandcoverMapping #sustainability
💻 TIMESTAMPS:
0:00 - Introduction to Land Cover Mapping
0:41 - Creating Land Cover Maps Explained
1:40 - Utilizing Earth Engine for Mapping
2:47 - Overview of Classification Algorithms
5:08 - Common Mapping Challenges
8:01 - Bypassing Earth Engine Cache Limit
14:58 - Implementing Random Sampling Techniques
17:50 - Guide to Stratified Sampling Methods
20:22 - Understanding Spatial Autocorrelation
22:50 - Constructing Feature Vectors
27:43 - Exploring Dynamic World Dataset
30:10 - Analyzing Texture in Land Cover
34:42 - Emma Izquierdo-Verdiguier on Ensemble Classifiers
44:34 - Latest Announcements in Land Cover Mapping
49:03 - Interactive Q&A Session
51:31 - Merging Classifiers Best Practices
53:47 - Conclusion and Final Thoughts
🎙️ SPEAKERS:
- Noel Gorelick, Google
- Emma Izquierdo-Verdiguier, University of Natural Resources and Life Sciences
