1. What contributions have the authors mentioned in the paper "Coverage optimized active learning for k - nn classifiers" ?
In this paper, the authors propose a batchmode active learning algorithm for efficient training of kNN classifiers, that substantially reduces the amount of training required.. The authors propose a coverage formulation that enforces selected samples to be distributed such that all data points have labeled samples at a bounded maximum distance, given the training budget, so that there are labeled neighbors in a small neighborhood of each point.. Further the authors employ uncertainty sampling along with coverage to incorporate model information and improve classification.. Finally, the authors employ locality sensitive hashing for fast retrieval of nearest neighbors during classification, which provides 1-2 orders of magnitude speedups thus allowing real-time classification with large datasets.. This work may not be copied or reproduced in whole or in part for any commercial purpose.. Permission to copy in whole or in part without payment of fee is granted for nonprofit educational and research purposes provided that all such whole or partial copies include the following: a notice that such copying is by permission of Mitsubishi Electric Research Laboratories, Inc. ; an acknowledgment of the authors and individual contributions to the work ; and all applicable portions of the copyright notice.
read more
2. What are the future works in "Coverage optimized active learning for k - nn classifiers" ?
Future work will focus on extending the work to ensemble classifiers, which perform very well on many real-world applications.
read more
3. What is the main goal of active learning?
Work in active learning initially focused on binary classification [3], [19]–[21], where it provided substantial reduction in label complexity (amount of annotated training data required) to achieve a certain classifier performance.
read more
4. What is the way to find nearest neighbors?
Given a dataset P and a query q, in the (R, c)–near neighbor (NN) problem [9], one has to retrieve points p such that d(p, q) ≤ cR, if there exists a point in P within distance R from q.
read more
