【在 W*W 的大作中提到】 : 比如 一个 model : y = ax+b : data 是加了noise的 x 和 y。 请问 怎么用MC 方法做fitting,例如再mablab里的算 : 法。
s*t
3 楼
since there's error in x too, better to define the cost as the sum of the distances between the points and the line. this problem is difficult to solve, but since he wants to use MC anyways ...