Hello,
I would appreciate if someone could advise how I can get the table where I can see the most important layers for RT classification in eCognition. I have around 20 layers as input variables, and I want to select only the most important ones and run the classification again.
Thank you!
https://docs.ecognition.com/v9.5.0/eCognition_documentation/User%20Guide%20Developer/6%20About%20Classification.htm
Hi,
You could run a search (or a ranking) method such as "best first" (or ranker method). Then, evaluate the merits of the selected subset using DT. Finally, apply DT using the best feature subset.
Note that all the highlighted process (feature selection + classification) must be evaluated at the same time using, e.g., ten-fold cross-validation.
HTH.
Dr. Samer Sarsam
Cristian Ramos-Vera , Thanks for the link, but I could not find the answer on my question there.
- Badges
- Science topic
- Similar topics
- Biological Science
- Kinesiology
- Running