Instance Selection for the Nearest Neighbor Classifier: Connecting the Performance to the Underlying Data Structure.
Vicente GarcíaJosé Salvador SánchezAlberto Ochoa OrtízAbraham López-NajeraPublished in: IbPRIA (1) (2019)
Keyphrases
- nearest neighbor classifier
- instance selection
- nearest neighbor
- data structure
- feature and instance selection
- knn
- index structure
- high dimensional data
- feature selection
- training samples
- distance measure
- training set
- supervised learning
- text classification
- data points
- multiple instance learning
- data reduction
- classification accuracy
- multi class
- distance function
- learning algorithm
- image processing
- decision trees
- dynamic time warping
- evaluation criteria
- concept drift
- classification rules
- knowledge discovery and data mining
- data analysis
- high dimensional
- semi supervised learning