Retraining: A Simple Way to Improve the Ensemble Accuracy of Deep Neural Networks for Image Classification.
Kaikai ZhaoTetsu MatsukawaEinoshin SuzukiPublished in: ICPR (2018)
Keyphrases
- neural network
- image classification
- high accuracy
- ensemble pruning
- pattern recognition
- feature extraction
- computational cost
- classification accuracy
- bag of words
- prediction accuracy
- random forest
- feed forward
- neural network model
- back propagation
- image representation
- sufficiently accurate
- neural network ensemble
- ensemble learning
- akaike information criterion
- multi label
- error rate
- image features
- training set
- genetic algorithm