基于多样本的在线支持向量回归算法

提出一种针对多样本的在线支持向量回归（SVR）算法，以解决目前SVR在线训练算法每次只能处理1个样本的问题．算法以拉格朗日乘数法和库恩一塔克（KKT）条件为基础，逐步改变样本的系数，并在每次迭代中保持原来的样本满足KKT奈件，最终使所有训练样本满足KKT条件．实验表明，该方法可有效更新SVR模型，且计算效率相比于基于单样本的在线回归算法有较大的优势．

On-line support vector regression with multiple samples

An on-line support vector regression（SVR） algorithm with multiple samples algorithm is proposed in order to overcome the drawbacks of the previously proposed algorithm which can only deal with one training sample at a time.The algorithm is based on the Lagrangian multiplier method and the KKT conditions.At each step of iteration,the algorithm modifies the Lagrangian multipliers of the updated samples while making sure the KKT conditions are fulfilled for all other training samples.The on-line training algorithm terminates when all training samples fulfill the KKT conditions.Experimental result shows that the proposed algorithm can update the SVR model on-line effectively and it is much more efficient than the on-line training algorithm with single samples.