Learning Representations for Neural Network-Based Classification Using the Information Bottleneck Principle.
Rana Ali AmjadBernhard C. GeigerPublished in: IEEE Trans. Pattern Anal. Mach. Intell. (2020)
Keyphrases
- neural network
- incremental learning
- learning process
- online learning
- learning algorithm
- supervised learning
- image classification
- dictionary learning
- learning phase
- classification accuracy
- pattern recognition
- data sets
- feature extraction
- decision trees
- cost sensitive
- automatic classification
- automatically discovering
- multi category
- discriminative learning
- classification systems
- training process
- learning problems
- machine learning methods
- machine learning algorithms
- support vector machine svm
- training samples
- prior knowledge