Local regularization for multiclass classification facing significant intraclass variations

We propose a new local learning scheme that is based on the principle of decisiveness: the learned classifier is expected to exhibit large variability in the direction of the test example. We show how this principle leads to optimization functions in which the regularization term is modified, rather than the empirical loss term as in most local learning schemes. We combine this local learning method with a Canonical Correlation Analysis based classification method, which is shown to be similar to multiclass LDA. Finally, we show that the classification function can be computed efficiently by reusing the results of previous computations. In a variety of experiments on new and existing data sets, we demonstrate the effectiveness of the CCA based classification method compared to SVM and Nearest Neighbor classifiers, and show that the newly proposed local learning method improves it even further, and outperforms conventional local learning schemes.