0

I am looking at enrolling in a university paper on modelling but am faced with an option of either a paper based on linear regression or logistic regression.

To be honest I don't quite know the difference. My limited understanding is that logistic regression relates more to probability more than the other. Again, I have a limited understanding so I may have misunderstood this.

Is any one able to shed some light on how they may differ and what real world applications may suit one over the other.

thanks

B.Miller
  • 669
  • 1
  • 7
  • 13
  • 3
    Linear regression involves a dependent variable that is continuous, whereas logistic regression involves a dependent variable that can only take 2 possible values (e.g., success and failure, and you model the probability of success). They're both important, but if you can only study one, I'd recommend linear regression (and even if you were going to study both, it's probably better to study linear regression first). – mark999 Jun 15 '15 at 04:08
  • 1
    If you don't know anything about linear regression, you won't be successful in studying logistic regression. In addition to what @mark999 said, there are also very different assumptions as to the structure of the data between linear and logistic regression, as well as how to interpret the output (i.e. log-odds). Linear regression is definitely the way to go in my opinion. For an example of what Linear regression can solve: predicting housing prices, number of students that pass an exam, temperature in July, etc... – ilanman Jun 17 '15 at 01:05
  • See https://stats.stackexchange.com/questions/27651/how-would-you-explain-generalized-linear-models-to-people-with-no-statistical-ba – kjetil b halvorsen Oct 07 '17 at 11:56

0 Answers0