paddockpass/ML/udemy/NOTES.md

SECTION 2

• you need to split the training set and a test set to balance the machine learning (you train on the test set and test those assumptions on the test set) ? what is categorical data, why whould you use it?

• feature scaling: put all values on the same scale so the larger number do not destroy other numbers => standardisation ou normalisation

Linear regression

Formula: y = b0 + b1*x1

y is the dependent variable. What is the value that changes in our model, the part that we want to understand from the work. How is this value changing.

x is the independant variable, the one that has an implied association with y.

b1 is the coef for the independant variable, how a unit change in x1.