Microalgae classification using semi-supervised and active learning based on Gaussian mixture models.
Paulo Drews Jr.Rafael G. ColaresPablo MachadoMatheus de FariaAmália DetoniVirgínia TavanoPublished in: J. Braz. Comput. Soc. (2013)
Keyphrases
- semi supervised
- gaussian mixture model
- active learning
- partially labeled data
- supervised learning
- feature vectors
- unsupervised learning
- mixture model
- semi supervised learning
- labeled data
- density estimation
- pattern recognition
- training set
- feature space
- selective sampling
- co training
- pattern classification
- classification accuracy
- speaker recognition
- em algorithm
- unlabeled data
- covariance matrices
- machine learning
- generative model
- feature extraction
- probability density function
- pairwise
- cost sensitive
- support vector
- image classification
- labeled examples
- class labels
- prior knowledge
- pairwise constraints
- pool based active learning
- probability density
- training examples
- text classification
- object recognition
- similarity measure
- image segmentation
- feature selection