最小二乘算法比较经典的著述，可以看一下，对学习LM，高斯牛顿法很有启示-Comparison of classical writings least squares algorithm, can look

BP神经网络代码，使用LM算法（高斯牛顿法的改进，具有很好的鲁棒性）-BP neural network code, use the LM algorithm (modified Gauss-Newton method has good robustness)

滴流的lbp算法 主要是层流的高度问题 我自己改造的层流重建 - 加权修正高斯牛顿 -算法算法-The trickle of LBP algorithm Mainly is the height of the laminar flow problem My own transformation of laminar flow reconstruction- weighted modified gauss-newton algorithm- algorithm