Distance Measures in Training Set Selection for Debt Value Prediction.
Tomasz KajdanowiczSlawomir PlamowskiPrzemyslaw KazienkoPublished in: PerMIn (2012)
Keyphrases
- distance measure
- training set
- euclidean distance
- similarity measure
- training data
- dynamic time warping
- distance function
- data sets
- bhattacharyya distance
- prediction error
- short term
- vector space
- kullback leibler
- nearest neighbor
- classification accuracy
- selection criterion
- histogram intersection
- edit distance
- training samples
- random walk
- nearest neighbor classification
- mutual information
- decision trees