0
$\begingroup$

I have a data set with ~80 records, with ~8 features. I want to predict one of the features in future records. The feature is numeric and discrete. It ranges between -30 up to 140 with steps of 5. Until now I wanted to predict another feature which is boolean, so I used logistic regression. Which method should I use here? Maybe some kind of particle filter?

Thanks!

$\endgroup$
1
  • $\begingroup$ I forgot to mention that this feature doesn't normally distributed $\endgroup$ – Noam Peled Nov 26 '11 at 23:44
2
$\begingroup$

Depending upon the distribution of your data you can re-apply the GLM here, as you did it with logistic regression, and choosing an appropriate link function.

$\endgroup$
0
1
$\begingroup$

Technically your problem is a multiclass classification problem not a regression problem. The difference is subtle, but if you expect the predicted value to behave like a regression variable, where 5.0 is "closer" to 6.0 than it is to 50.0, then you can just use any regression technique and round the result to the nearest integer.

Particle filters are an inference technique for sequential distributions, I don't see how they are related at all.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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