Effects of transfer learning for handwritten digit classification in a small training sample size situation.
Yoshihiro MitaniNaoki YamaguchiYusuke FujitaYoshihiko HamamotoPublished in: AICCC (2022)
Keyphrases
- sample size
- transfer learning
- handwritten digits
- small sample
- small sample size
- model selection
- text classification
- structure learning
- machine learning
- learning tasks
- supervised learning
- labeled data
- cross domain
- upper bound
- training set
- pac learning
- cross domain learning
- handwritten chinese
- training samples
- semi supervised learning
- transfer knowledge
- machine learning algorithms
- active learning
- progressive sampling
- collaborative filtering
- cross validation
- learning algorithm
- class labels
- support vector machine
- decision trees
- feature selection
- character recognition
- image classification
- worst case
- feature space
- feature extraction
- training examples
- text categorization
- unlabeled data
- multi task
- text mining
- feature vectors
- support vector