Basic intuition of Logistic Regression

Basic intuition of Logistic Regression


Hi all, Today’s video helps build an intuition
around a classification algorithm called logistic regression. We won’t be using any formulas in the given
example. So without wasting for the time, let’s get
started. Climbing a hill has always been a big challenge. However a genius scientist created a special
vehicle, which acts like an escalator that will move you up the hill to complete the
challenge. This is how the vehicle looks The vehicle has a scale which looks for a certain weight and height combination The closer you are
to that weight & height combination the further the machine will take you up the hill. In order for someone to use the machine a person must step onto the scale of the machine and calculate
and the machine would calculate his height and weight. If the special machine takes someone at least
50 percent of the hills height, then you can easily say that if climbed the hill which
is positive class and their classified as completed the challenge. Higher the machine
takes them the more confident you are that they’ve completed the challenge. So let’s look at the first example. I have my friend John here who easily surpasses
the 50% barrier. Thus he has completed the challenge. However, my second friend. Jack the machine couldn’t reach the 50% threshold
mark thus he failed to complete the challenge. The vehicle that the scientist designed is inherently
using a logistic regression model. Some weights multiplied by features which
in our case are made of the person as well as the height of the person squashed by a sigmoid
function helps us reach the prediction of the classes. Calculation of the right weights to be
multiplied by features Is calculated using something called as gradient descent. Hope you liked the simple intuition based
example for logistic regression do like and subscribe to my channel. Thank you so much.

One comment

  1. Please give more videos on detailed explanation of logistic regression ,how to calculate weights and error?

Leave a Reply

Your email address will not be published. Required fields are marked *