Which one is better method for dimensinality reduction PCA, Forward elimination or backward elimination?

machine_learning
predictive_model
dimensionality

#1

Hi Friends,

Recently, I was working on data science problem having ~200 features and 30,000 records in train data set so it is difficult to perform univariate and bivariate analysis and identify most significant variables. I have gone through the dimensionality reduction techniques like PCA, Forward and Backward elimination and others. Can you help me to understand which method delivers better result compare to others?

Thanks,
Imran