An ADMM-based approach for multi-class recursive parameter estimation.
Valentina BreschiAlberto BemporadIlya V. KolmanovskyPublished in: CDC (2021)
Keyphrases
- parameter estimation
- multi class
- model selection
- maximum likelihood
- pairwise
- support vector machine
- least squares
- markov random field
- feature selection
- em algorithm
- multiple classes
- object detection
- binary classification
- multi class classification
- expectation maximization
- parameter estimation algorithm
- cost sensitive
- parameter estimates
- error correcting output codes
- multi class boosting
- total variation
- multi class problems
- multi class object detection
- binary classifiers
- image segmentation
- base classifiers
- convex optimization
- semi supervised
- learning algorithm
- data sets