Linear regression is a widely used ML algorithm. So far I have only encountered 'boring' applications of it. (e.g predict sales for next quarter, predict housing prices for next year , predict population of a country by 2020 etc.)

What are some interesting/cool application of linear regression? Finding cool applications greatly helps in motivating myself to learn, hence the quest.

Cool as in: Application to stock trading, application to video games, application to astronomy, application to sports betting, application to airfare prediction

  • $\begingroup$ What is cool or boring for you ? $\endgroup$ Jul 20, 2015 at 16:02

2 Answers 2


The way that you have phrased this question makes it tough for people to answer without first offering you some background on linear regression (LR). Its great that you are interested in learning some ML and LR is a great place to start.

Linear regression is really just finding a line (or plain or hyperplain) that maps a relationship between two or more variables or features. The important requirement is that the target variable be a continuous numerical value. So its less about finding a problem for which linear regression "works" and more about finding some data that interests you and playing with it using linear regression.

I suggest you download some open data in the realms that you have mentioned. There is plenty of open data in every topic that you mentioned that contain continuous numerical values that you can predict using other features of the data.

  1. stock trading
  2. video games
  3. astronomy
  4. sports betting
  5. flight time prediction

Also think about taking some sort of online MOOC as this will help you gain some footing in the subject. Andrew Ng's Coursera on Machine Learning is highly recommended as a starting point and include some linear regression during the first portion and scikit-learn is a great Python based library.

  • $\begingroup$ Thanks. I was thinking of R. Any reason for python over R? $\endgroup$
    – Victor
    Jul 20, 2015 at 17:42
  • 1
    $\begingroup$ No specific reason for python. R vs. Python is a hotly debated topic. R has been around for a long time so has tons of packages. Python is more user friendly and is the future of data science. Python will make you happier and will help you get things done faster. If you really need R then you can call R from within Python. Someone will probably respond and tell you why R is best, but I like python. Take a look at all of the following: pandas, scikit-learn, numpy, matplotlib, ipython notebook. There's a great book "Python for Data Analysis" and the scikit learn user guide is fantastic. $\endgroup$
    – AN6U5
    Jul 20, 2015 at 17:50
  • $\begingroup$ And I thought Julia was the future of data science ;) $\endgroup$ Jul 21, 2015 at 12:44

I think those were the coolest. I recommend you that video also https://www.khanacademy.org/math/probability/regression/regression-correlation/v/regression-line-example


Not the answer you're looking for? Browse other questions tagged or ask your own question.