Best feature selection method


#1

Hi,
I am new to the Analytics field.I have a dataset of 76 variables and 1059 observations. The dependent variable is continuous and the independent variables are continuous and categorical in nature.which will be best feature selection method for such type of problem?

Thank you,


#2

Too generic a question as there are variety of methods to explore. Considering the number of observations and variables(predictors) there are multiple options.

Easiest to start with is running regression and finding significance and eliminating variables.

PCA based regression is another option where you can choose the right number of PC’s to run a machine learning algorithm.

No matter what you choose you have to spend some time to look at predicting performance on the test sample and decide.