Optimum Design Parameters of the Classifiers for Omni-Font Machine-Printed Numeral Recognition Based on the Minimum Classification Error Criterion.
Katsumi MarukawaKazuki NakashimaHiroshi ShinjoYoshihiro ShimaHiromichi FujisawaPublished in: MVA (1998)
Keyphrases
- minimum classification error
- statistical pattern recognition
- maximum likelihood estimation
- optimum design
- character recognition
- classification error
- gaussian mixture
- discriminative training
- hidden markov models
- multi objective
- pattern recognition
- handwritten numerals
- handwritten documents
- feature selection
- em algorithm
- maximum likelihood
- gibbs sampling
- feature extraction
- expectation maximization
- supervised learning
- training set
- mixture model
- printed documents
- class labels
- text mining
- image analysis
- decision trees