An improvement on parametric $$\nu $$ ν -support vector algorithm for classification.
Saeed KetabchiHossein MoosaeiMohamad RazzaghiPanos M. PardalosPublished in: Ann. Oper. Res. (2019)
Keyphrases
- support vector
- classification algorithm
- dynamic programming
- multi class classification
- computational complexity
- learning algorithm
- detection algorithm
- preprocessing
- machine learning
- matching algorithm
- optimal solution
- support vector machine svm
- benchmark data sets
- classification accuracy
- similarity measure
- classification scheme
- objective function
- accuracy rate
- np hard
- recognition algorithm
- search space
- worst case
- expectation maximization
- cross validation
- generalization ability
- kernel logistic regression
- svm classifier
- segmentation algorithm
- particle swarm optimization
- simulated annealing
- multi class
- cost function
- significant improvement
- training set
- pattern recognition
- training data
- decision trees