Divide-and-conquer ensemble self-training method based on probability difference.
Tingting LiJia LuPublished in: J. Ambient Intell. Humaniz. Comput. (2022)
Keyphrases
- preprocessing
- cost function
- high precision
- significant improvement
- objective function
- cross entropy
- synthetic data
- clustering method
- support vector machine svm
- high accuracy
- semi supervised
- d objects
- model selection
- unsupervised learning
- classification accuracy
- detection algorithm
- probabilistic model
- detection method
- prediction accuracy
- artificial neural networks
- classification method
- test data
- pairwise
- data sets