Analysis of new techniques to obtain quality training sets

174 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Learning kernel logistic regression in the presence of class label noise (2014, background): “The seemingly straightforward approach is by means of data preprocessing where any suspect samples are removed or relabelled [7,1,29,37,31,18].”
On the characterization of noise filters for self-training semi-supervised in nearest neighbor classification (2014, background): “Nearest Centroid Neighbor Edition (NCNEdit) [46]: This algorithm defines the neighborhood, taking into account not only the proximity of prototypes to a given example, but also their symmetrical distribution around it.”
Understanding classifier errors by examining influential neighbors (2015, methodology): “In [24], the training data set is cleaned by automatically removing any training example with many examples with the opposite label nearby.”
Editing Training Data for kNN Classifiers with Neural Network Ensemble (2004, methodology): “The Depuration algorithm was regarded as the first prototype selection approach [9], which consists of removing some “suspicious” training examples while changing the class labels of some other examples.”