Paper Info
Title
Distinguishing Mislabeled Data from Correctly Labeled Data in Classifier Design
Abstract
We have developed a method for distinguishing between correctly labeled and mislabeled data sampled from video sequences and used in the construction of a facial expression recognition classifier. The novelty of our approach lies in training a single, optimal classifier type (a Support Vector Machine, or SVM) on multiple representations of the data, involving different "discriminating" subspaces. Results of a preliminary study on the discrimination of "high stress" vs. "low stress" facial expression data by this method confirms that our novel approach is able to distinguish subproblems where labeling is highly reliable from those where mislabeling can lead to high error rates. In helping detect data sub-samples which yield misleading classification results, the method is also a rapid, highly efficient cross-validated approach for eliminating outliers.
Year
DOI
Venue
2004
10.1109/ICTAI.2004.52
ICTAI
Keywords
Field
DocType
classifier design,novel approach,distinguishing mislabeled data,high stress,correctly labeled data,optimal classifier type,facial expression data,mislabeled data,data sub-samples,low stress,facial expression recognition classifier,high error rate,efficient cross-validated approach,image classification,facial expression,support vector machine,learning artificial intelligence,face recognition,support vector machines,gesture recognition,cross validation,error rate
Structured support vector machine,Facial recognition system,Pattern recognition,Computer science,Support vector machine,Outlier,Artificial intelligence,Relevance vector machine,Margin classifier,Classifier (linguistics),Contextual image classification,Machine learning
Conference
ISSN
ISBN
Citations 
1082-3409
0-7695-2236-X
10
PageRank 
References 
Authors
0.72
13
5
Name
Order
Citations
PageRank
Sundara Venkataraman1100.72
Dimitris N. Metaxas28834952.25
Dmitriy Fradkin334419.25
casimir a kulikowski4616299.37
Ilya Muchnik532347.03