Programmatically Accessing the Classification Learner
6 Ansichten (letzte 30 Tage)
Ältere Kommentare anzeigen
Jacob Mevorach
am 12 Okt. 2017
Kommentiert: Jacob Mevorach
am 2 Nov. 2017
Is there anyway to programmatically access the classification learner and have it train all the models based on a set of classification data without having to actually open it from the app menu? I'm interested in getting the percent accuracy numbers associated with each model. If there was a way to do this without having to manually open the classification learner it would be very helpful.
0 Kommentare
Akzeptierte Antwort
Rajesh Balagam
am 17 Okt. 2017
It's not currently possible to programmatically access the results from the Classification Learner.
At present, only way to get those values is to generate and run code for the desired models. Even in this case the accuracy values may be slightly different from the app due to randomness in choosing the training and validation set.
You can try setting the random seed before using the app, and then use the same random seed when running the generated code. However, the exact match for the results is not guaranteed.
Weitere Antworten (0)
Siehe auch
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!