Rigorous proof of termination of SMO algorithm for support vector Machines.
Norikazu TakahashiTetsuo NishiPublished in: IEEE Trans. Neural Networks (2005)
Keyphrases
- experimental evaluation
- preprocessing
- optimization algorithm
- computational cost
- support vector
- computational complexity
- recognition algorithm
- significant improvement
- cost function
- np hard
- improved algorithm
- times faster
- optimal solution
- dynamic programming
- matching algorithm
- decomposition algorithm
- learning algorithm
- decomposition method
- generalization ability
- neural network
- provably correct
- cross validation
- detection algorithm
- particle swarm optimization
- linear programming
- simulated annealing
- worst case
- k means
- artificial neural networks