[Solved] how to use svm classifier in feature extraction

Question

This question is about how to use these matlab functions:

If you are trying to classify entire videos you will have one label per video, i.e. assign 1 or 0 to “single block of 40 by 5 like 20 rows”, In which case your Training data matrix should be 20×200 (20 videos with 200 features) and you will have a 20×1 Group label vector.

If you are trying to classify individual frames in each video your Training data matrix should be 800×5 (800 frames of 5 features) and you will have a 800×1 Group label vector.

According to classperf documentation:
“classout must contain the same number of elements as truelabels.”
It also has a good example of how to use classperf.

Accepted Answer

This question is about how to use these matlab functions:

If you are trying to classify entire videos you will have one label per video, i.e. assign 1 or 0 to “single block of 40 by 5 like 20 rows”, In which case your Training data matrix should be 20×200 (20 videos with 200 features) and you will have a 20×1 Group label vector.

If you are trying to classify individual frames in each video your Training data matrix should be 800×5 (800 frames of 5 features) and you will have a 800×1 Group label vector.

According to classperf documentation:
“classout must contain the same number of elements as truelabels.”
It also has a good example of how to use classperf.