A fast approximation algorithm for 1-norm SVM with squared loss.
Li ZhangWeida ZhouZhao ZhangJiwen YangPublished in: IJCNN (2015)
Keyphrases
- learning algorithm
- support vector machine svm
- classification algorithm
- recognition algorithm
- multi class classification
- dynamic programming
- optimization algorithm
- computational cost
- cost function
- k means
- training data
- optimal solution
- clustering algorithm
- segmentation algorithm
- svm classifier
- particle swarm optimization
- objective function
- markov random field
- search space
- multi class
- support vectors
- approximation ratio
- preprocessing
- training algorithm
- detection algorithm
- decision function
- polygonal approximation
- theoretical analysis
- expectation maximization
- simulated annealing
- worst case
- semi supervised
- probabilistic model
- feature vectors
- computational complexity
- search algorithm
- support vector