You are here: Home > Conclusions

Conclusions

From this research, a number of conclusions can be made on the application of Support Vector Machines to image data:

  • Support Vector Machines can perform well on large datasets, such as the classification of every pixel in an image.
  • Reducing the overlap between the different classes of the training and test data reduces the training time of the Support Vector Machine and increases the generalisation performance of the classifier to test data.
  • Reducing the background variation within the training and testing examples allows the size of the training dataset to be reduced without affecting the generalisation performance of the trained classifier.