标签:csdn nbsp 模型 .net article tail 公式 线性 lis
?BA/?A = BT
?ATB/?A = B
?ATBA/?A = 2BA
hθ(x) = xθ
J(θ) = 1/2(Xθ-Y)2
令偏导?J(θ)/?θ等于0,得:
θ = (XTX)-1XTY
J(θ) = 1/2(Xθ-Y)2+1/2λθ2
令偏导?J(θ)/?θ等于0,得:
θ = (XTX+λI)-1XTY
参考文献:
https://blog.csdn.net/fleurdalis/article/details/54931721
https://blog.csdn.net/u013363719/article/details/22752893
标签:csdn nbsp 模型 .net article tail 公式 线性 lis
原文地址:https://www.cnblogs.com/jhc888007/p/9501258.html