I can’t grasp the logic behind logistic regression..
Can you help me?
Great to have you in the course and thanks for reaching out!
Nothing to worry about. This is a topic that requires not only solid theoretical understanding, but also a certain amount of practice.
Theoretically, as any other regression analysis, logistic regression aims to explain the relationship between the independent variables and the dependent variable. With one important distinction – the dependent variable must be binary (i.e. 1 or 0; ‘yes’ or ‘no’, and others).
Therefore, while dealing with logistic regression, albeit you are dealing with a regression, there are several techniques that are peculiar just for this type.
Please refer to Section 5, Lecture 15, Minute 3:20 for a theoretical explanation.
However, if you are more interested in the subject, I would like to suggest to subscribe to our course The Data Science Course 2018: Complete Data Science Bootcamp. There we apply this topic in an exercise, and discuss it in detail.
Hope this helps.