SVM SUCCEEDS
- SVM achieves an error rate of 32.1% using
RGB + Long Lines as features which is the first time we
have achieved a less error performance when compared to that of
intelligent guessing using any classification algorithm.
- No confusion between the LSBE and HSBE class.
- Shown below is the confusion matrix that was obtained as a
result of testing on set 01 using rgb and long lines as
feature vectors. The error rate for this experiment is 32.1%.