量子粒子群优化算法在训练支持向量机中的应用

训练支持向量机的本质问题就是求解二次规划问题，但对大规模的训练样本来说，求解二次规划问题困难很大。遗传算法和粒子群算法等智能搜索技术可以在较少的时间开销内给出问题的近似解。量子粒子群优化（QPSO）算法是在经典的微粒群算法的基础上所提出的一种有较高收敛性和稳定性的进化算法。将操作简单而收敛快速的QPSO算法运用于训练支持向量机，优化求解二次规划问题．为解决大规模的二次规划问题开辟了一条新的途径。

Application of quantum-behaved particle swarm optimization in training support vector machines

The key problem of training support vector machines is how to solve quadratic programming problem, but for large training examples, the problem is too difficult. The intelligent search techniques, such as genetic algorithms and particle swarm optimization algorithm, can reach a similar solution of problem in less time. Quantum-behaved particle swarm optimization (QPSO) developed on the basis of classical particle swarm optimization is a method with better convergence and stability. Simple and rapid QPSO algorithm is applied in training support vector machines to solve quadratic programming problem. It is a new way for solving quadratic programming problem with a large number of example vectors.