Archive for October, 2009

How to Make Youself Understood

Thursday, October 29th, 2009

DK: I tried to write things up in a way that was jargon-free, so the nonspecialist would understand it. What I succeeded in doing is making it so that the specialist can understand it, but if I hadn’t tried to write jargon-free, then I would have written for specialists, and then the specialists wouldn’t be able to get it either.

写给普通人看,专家能看懂;写给专家看,谁也看不懂。高德纳如是说。

Short Note on Linear Regression Implementations

Monday, October 5th, 2009

Linear regression和相应的ridge regression是简单有效的统计模型。一般的实现都会吧feature normalize以后再计算相应的系数,得到结果以后可以再把结果换算成没有normalize之前的值。对于linear regression,怎么做关系不大。但是对于ridge regression这样和没有normalize直接计算weight得到的结果是不同的。

下面是引自statistical learning里面的话。

The ridge solutions are not equivalent under scaling of the inputs, and so one normally standardizes the inputs before solving…

很多ridge regression的实现,有unscale的选项,但是它们的实际实现通常是把feature normalize,然后把计算得到的结果再除以之前normalize时用的scaling factor。这样的结果和完全不scale是不一样的。这是不是真的unscale呢?我认为有误导的嫌疑。