I have a continuous dependent variable. 5 categorical independent variable with 7-12 levels in each. Converting into dummy variables and using regression doesn't sound good as there will be so many variables. Which analytical methods can be used here ?
$\begingroup$
$\endgroup$
5
asked Apr 15, 2016 at 6:50
-
$\begingroup$ Is it a ' self study" question? If yes give some data and indicate the objective. $\endgroup$– user10619Apr 28, 2016 at 12:22
-
2$\begingroup$ You really need to give us more details! How many observations, for instance. Or better: tell us about the applied problem, in the language of the application. Then maybe ... $\endgroup$– kjetil b halvorsen ♦Sep 20, 2016 at 16:22
-
$\begingroup$ 9053166 2330 144 21 196063 196047 129639 10316 1817 2310 308 13 1 4 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 1 0 1 0 0 0 1 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 1 1 0 1 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 $\endgroup$– Arpit SisodiaSep 22, 2016 at 5:10
-
$\begingroup$ this is just one observation. All predictors are categorical, Dependent variable is click probability.We have more than 1 million observations. $\endgroup$– Arpit SisodiaSep 22, 2016 at 5:14
-
$\begingroup$ Unless there is some kind of structure in the categorical predictors (maybe they are ordinal, or there is some kind of random effects you can exploit) I am afraid it doesn't look like a soluble problem. You essentially have about $10^5 = 100000$ unique groups and one million observations, thus roughly, your MLE estimate is going to be the average value from 10 observations. $\endgroup$– AlexAug 24, 2018 at 6:43
Add a comment
|
1 Answer
$\begingroup$
$\endgroup$
ANOVA is recommended when you have a continuous dependent variable, and a categorical independent variable.
Logistic regression could be used if you convert the dependent variable into a categorical one.
