// copyright by Peter Huber, Singapore, 2006 // this code is provided as is, bugs are probable, free for any use at own risk, no // responsibility accepted. All rights, title and interest in and to the accompanying content retained.  :-)

Huber损失最小化学习和高斯核模型 的Huber损失最小化，竞赛用到过，效果不错(Huber loss minimization learning and Gauss kernel model Huber loss minimization, the competition used, the effect is good)

包含基于Huber函数和最大相关熵的抗差滤波算法，以及与卡尔曼滤波算法的对比，用于GNSS导航定位粗差处理。同时主函数中包含混合高斯随机数生成方法。(The Huber function and maximum correntropy criterion based robust Kalman filter is included, which can be used for biases in measurements for GNSS positioning. The mixed Gauss