Multi-Classifier Adaptive Training: Specialising an Activity Recognition Classifier Using Semi-supervised Learning.
Bozidara CvetkovicBostjan KaluzaMitja LustrekMatjaz GamsPublished in: AmI (2012)
Keyphrases
- semi supervised learning
- activity recognition
- labeled training data
- unlabeled data
- training data
- co training
- labeled data
- supervised learning
- semi supervised
- training set
- training examples
- labeled and unlabeled data
- transfer learning
- labeled instances
- training samples
- training process
- active learning
- multiple kernel learning
- unlabeled samples
- human activity recognition
- text classification
- class labels
- support vector machine
- labeled examples
- human activities
- feature selection
- semi supervised classification
- training and test data
- semi supervised learning methods
- data sets
- manifold regularization
- multiple instance learning
- action recognition
- learning algorithm
- machine learning
- text categorization
- label propagation
- training instances
- unsupervised learning
- multi view
- nearest neighbor
- knowledge discovery
- prior knowledge
- learning environment
- unlabeled instances