Landslide Susceptibility Mapping using Machine Learning (Project 228)

[robertogl]

Contribute to the discussion by asking and/or answering questions, commenting, or sharing your ideas for solutions to project #228

[Emma2020-star]

Hell, how do l access the training data

[mw-pc]

Hello, section #2 of reference #1 points to the following https://github.com/jeffreyevans/GradientMetrics as a means in which the authors obtained the necessary data e.g., Distance from roads, distance from geological faults, distance from water streams, land use type, geology type, etc.

[ankur18439]

Hi @Emma2020-star! I think we'll have to generate the input data ourselves. Some organizations provide landslide inventories for data labeling (I remember NASA has a publicly available landslide inventory that you can use). The other datasets will have to be processed using digital elevation models and satellite imagery.

[acetarik]

Hey! How do I access the dataset for training my nn?

[Emma2020-star]

hello, you can navigate to the various portals that track landslide occurrences, they have a page for their datasets.

[acetarik]

Thanks! It would be helpful if you could provide me with a link, it'll let me know if i am on the right page or not :3

[mw-pc]

Hello, consider search IEEE Explore. Here's an example search result.

[ankur18439]

Hey @acetarik! It's not a one-stop solution - you'll have to compile it from at least 2-3 data sources. You can digitize your labeling data (the landslide locations) in GIS but it is too much of a hassle sadly.