Data quantity and quality are crucial factors for data-driven learningmethods . In some target problem domains, there are not many data samples available, which could significantly hinder the learning process . We propose a weak adaptationlearning (WAL) approach that leverages unlabeled data from a similar sourcedomain, a low-cost weak annotator that produces labels based on task-specificheuristics, labeling rules, or other methods (albeit with inaccuracy), and asmall amount of labeled data in the target domain . Our experiments demonstratethe effectiveness of our approach in learning an accurate classifier with limited labeled data . Our approach first conductsa theoretical analysis on the error bound of the trained classifier and then introduces a multi-stage weak adaptation learning method to learn anaccurate classifier by lowering the errorbound. Our experiments demonstrate the effectiveness of the approach

Author(s) : Shichao Xu, Lixu Wang, Yixuan Wang, Qi Zhu

Links : PDF - Abstract

Code :

Keywords : data - weak - learning - approach - classifier -

Leave a Reply

Your email address will not be published. Required fields are marked *