A class-dependent weighted dissimilarity measure for nearest neighbor classification problems.
Roberto ParedesEnrique VidalPublished in: Pattern Recognit. Lett. (2000)
Keyphrases
- dissimilarity measure
- class dependent
- nearest neighbor
- distance function
- k nearest neighbor
- knn
- similarity measure
- high dimensional data
- maximum margin
- distance measure
- clustering method
- cost sensitive
- high dimensional
- feature space
- handwriting recognition
- locality preserving projections
- correlation filters
- euclidean distance
- distance metric
- edit distance
- data points
- feature ranking
- data analysis
- word alignment
- kernel trick
- data sets
- support vector
- pattern recognition
- pairwise