A study of the effect of feature reduction via statistically significant pixel selection on fruit object representation, classification, and machine learning prediction.
P. BeaulieuDalila B. MegherbiPublished in: CIVEMSA (2014)
Keyphrases
- statistically significant
- machine learning
- feature reduction
- pattern recognition
- machine learning methods
- decision trees
- feature selection
- support vector machine
- object representation
- data sets
- supervised learning
- text classification
- classification accuracy
- active learning
- hidden markov models
- feature extraction
- input image
- image classification
- neural network
- reinforcement learning
- data mining
- shape descriptors
- learning algorithm
- image processing
- classification algorithm
- prediction accuracy
- text categorization
- training samples
- rough sets
- co occurrence
- feature vectors