GeoAI: Machine Learning meets ArcGIS

[@lethargic_cow]

Proud of you Omar

[@gnash7977]

where can I see the road digitization? that looks interesting!

[@Matt-jm3ut]

kzbin.info/www/bejne/eZ7CoHSXhruKkJo skip to 9:00

[@amiyabasak7096]

sir, plz provide python code for that specific work

[@affansyed6000]

I find this video very interesting. I am a beginner to this field and I wanted to know a pathway to enter this field of geoAI. Can any one help me or suggest any courses. It would be great if someone drop their email ID and so that I can contact them. Thanks

[@Bixbeat]

So how exactly does ArcGIS Pro factor in in all of this? "We used many tools to preprocess the data and extract the features" - okay sure, but that doesn't tell me why Pro is better than any other solution. I've used FOSS solutions to extract features from geodata with reasonable comfort, why then would it be 'extremely challenging' to do it in anything other than ArcGIS Pro? This is the exact part where the speaker was supposed to sell ArcGIS Pro, but it ended up as an empty remark. Dissapointing to say the least. This presentation really doesn't sell me on the idea that GeoAI is anything more than typical geoprocessing ahead of a regular pipeline.

[@EsriEvents]

Hi Alex! Thanks for watching. To answer your question, ArcGIS Pro along with ArcPy scripts provides streamlined geoprocessing with machine learning. While you can tackle the same large-scale projects with FOSS you’ll likely have to stand up and manage a database, prepare the data, form a spatial index, and take additional steps to arrive at your analysis. The hundreds of geoprocessing tools built into ArcGIS Pro (such as Near, SpatialJoin, CalculateField, Project, DeleteField, AddField, etc which we used in this demo), make preparing the training data set possible in an accurate and timely manner. Additionally, there are ArcGIS tools that don’t have direct FOSS substitutes - such as using EBK regression to interpolate weather feeds to get accurate weather data per accident, taking geography into consideration. We hope that helps!

[@oelakad]

very interesting

[@fageh100]

very interesting