A new method to determine basic probability assignment from training data.
Peida XuYong DengXiaoyan SuSankaran MahadevanPublished in: Knowl. Based Syst. (2013)
Keyphrases
- test data
- training data
- prior knowledge
- cost function
- significant improvement
- detection method
- synthetic data
- fully automatic
- preprocessing
- training set
- computational cost
- training samples
- high accuracy
- classification accuracy
- test set
- experimental evaluation
- similarity measure
- optimization method
- segmentation method
- training process
- data sets
- cross entropy
- semi supervised
- knn
- probabilistic model
- objective function
- multiscale
- image segmentation
- learning algorithm
- genetic algorithm
- machine learning