IPMix: Label-Preserving Data Augmentation Method for Training Robust Classifiers.
Zhenglin HuangXianan BaoNa ZhangQingqi ZhangXiaomei TuBiao WuXi YangPublished in: CoRR (2023)
Keyphrases
- training samples
- synthetic data
- input data
- training data
- test data
- missing data
- classification method
- prior knowledge
- data sets
- computationally efficient
- similarity measure
- data analysis
- training process
- data sources
- statistical methods
- training set
- data structure
- training phase
- noisy data
- missing values
- class membership
- class labels
- clustering method
- multi class
- support vector machine
- support vector machine svm
- supervised learning
- extracted features
- multi label classification
- label information