Five minutes of machine learning every day: how to use matrix to represent the sample data of multiple characteristic variables?

This paper mainly

We've learned about univariate linear regression , For example, the prediction of house prices , Only the size of the house , In fact, there is not only one feature in our machine learning problem , For example, the house price is also related to the number of floors and rooms , We now have a model with multiple variables , The features in the model are (x1,x2,…,xn). So how should we deal with this problem at this time ？

Linear regression , The learned model is not necessarily a straight line , The model is a straight line only when it is univariate , When there are many features, the model learned is hyperplane .

House prices

Let's take the housing price as an example , But now there are many characteristic variables .

m Represents the number of samples , There is also a symbol to be added here n, It represents the characteristic number .

x(i) On behalf of the i Training samples , Is the second in the characteristic matrix i That's ok , It's a vector . For example, our second sample in , We can express it as ：

xj(i) Represents the... In the characteristic matrix i OK, No