Prediction and Inference

In order to build a good foundation for Data Science it is really important that you have your math/stats basics clear.

Starting with Stats, Prediction and Inference are two most important terms. Why?? Let me explain with an example :-

A Rugby match(that’s all I could think of) Scotland v/s England. All fans are so tensed and wishing their favorite team to perform well and win the match. When you see these matches on T.V or live, you see people talking about past matches and how one team performed well. This is nothing but talking about data “Previously collected Data”. The data might store factors — location, weather, year etc, using this data and their experience the commentator will assume who might win the match. This is called “Prediction”. In statistics/analytics this work is done by some algorithm. They use collected data and predict the outcome of future. So, Who will win the Next Rugby match???

Following the similar example, even though we have past data to predict the future outcome, there are chances that prediction won’t work because of the changes in the factors mentioned above. Like change in weather condition.. Imagine having Rugby match in India in 42 temp.. Both England and Scotland won’t be able to play.. Too hot in India for them.. Jokes apart.. So, finding out which factor affects the outcome and relationship between factors and output, is called “Inference”..

For more in-depth and mathematical explanation please refer to — “An Introduction to Statistical Learning”