In this paper, a special support vector regression machine algorithm is proposed, within which smoothing function and method to solve LC1 type functions are combined to solve Newton-type algorithm.
分别采用光滑化函数法和求解lc1函数类型方法对牛顿型算法进行研究求解。
A cost function is presented, and by applying Gaussian-Newton type recursive prediction error based method, a stable and efficient online frequency estimation algorithm is derived.
在此基础上，建立了最小方差损失函数，并结合高斯·牛顿预测误差方法，提出了稳定的，高性能的，在线的复频率直接估计算法。
A cost function is presented, and by applying Gaussian-Newton type recursive prediction error based method, a stable and efficient online frequency estimation algorithm is derived.
在此基础上，建立了最小方差损失函数，并结合高斯·牛顿预测误差方法，提出了稳定的，高性能的，在线的复频率直接估计算法。
应用推荐