1

Though I'm a while in this field I recognized that I can't say for sure that I understood this basic, very simple equation $ \hat{y} = xw + w_0 $

I know $ w_0 $ denotes the bias term (mostly given as $ b $) which is basically the offset of the function. But how do I understand $ xw $? Does $ x $ represent the feature(s) like speed and acceleration so it would have two dimensions with $ x_i , i=0,1 $ Or does it stand for all the data points of a feature with $ x_i, i=0,..,n $ ? And $ w $ is then exactly, what?

Ben
  • 2,032
  • 3
  • 14
  • 23
  • Does this answer your question? [How to tell the difference between linear and non-linear regression models?](https://stats.stackexchange.com/questions/148638/how-to-tell-the-difference-between-linear-and-non-linear-regression-models) – mhdadk Jun 05 '21 at 19:07
  • I'm trying to grasp it, it's quite formal. But for X is "X is a vector of numbers" given. So this is not the feature, then but the datapoints itself? – Ben Jun 05 '21 at 22:04
  • Answers to this are in evidence in several thousand posts here on CV. Good search terms are [normal equations](https://stats.stackexchange.com/search?tab=votes&q=normal%20equations) and [design matrix](https://stats.stackexchange.com/search?q=design+matrix). Alternatively, search for [multiple regression interpretation](https://stats.stackexchange.com/search?tab=votes&q=multiple%20regression%20interpretation). – whuber Jun 06 '21 at 13:54
  • Thanks, but I couldn't see a post. Yes, there are a lot of posts which include this but this would mean investing hours of trying to understand all these totally different questions with varying notations and background motivations. I will try finding something on google, instead. – Ben Jun 06 '21 at 16:19
  • In case someone is stumbling over this question: https://towardsdatascience.com/weighted-linear-regression-2ef23b12a6d7 – Ben Jun 06 '21 at 16:52

0 Answers0