How to improve your model score with polynomial features? #machinelearning #coding #python #aiml

Añadido: 2026-05-17

[00:00:00]So guys, so far we have seen the linear regression model, how actually we can train the model and predict the data.

[00:00:04]But here is the limitation of linear regression model. So if I run the code, you can see the model score is just 0.69. So there is nothing wrong with this model, rather we have the data like this. So but during the extreme heat, our ice cream sales is going down. So that is not forming the linear graph, okay? So in that case, there comes the polynomial regression. You can import polynomial features, okay? We can create a instance for polynomial features, all right? And we can give the degree is equal to two. Two is for curve, one is for a straight line, okay? Let's take a variable, we can call it as temp poly.

[00:00:37]So we are going to transform the data, okay? So poly.fit transform, okay? And then we can pass the input data, which is nothing but the temperature, okay?

[00:00:46]Now we are going to use the same model, but instead of training that with temperature, we are going to train it with temp poly, the variable we have created, okay? And while evaluating the model also, let's replace the temp with temp poly. Now if I run the code, you should be seeing the model score has been improved.

[00:01:00]So here you go, it has been improved to 0.90.