A hybrid approach based on transfer and ensemble learning for improving performances of deep learning models on small datasets.
Tunç GültekinAybars UgurPublished in: Turkish J. Electr. Eng. Comput. Sci. (2021)
Keyphrases
- ensemble learning
- learning models
- learning algorithm
- generalization ability
- ensemble methods
- loss function
- learning tasks
- machine learning
- random forest
- semi supervised learning
- conditional random fields
- transfer learning
- machine learning algorithms
- benchmark datasets
- learning problems
- base classifiers
- feature selection
- concept drift
- classification models
- unlabeled data
- data sets
- image classification
- object recognition
- image processing