On a daily basis, I build predictive models (namely, logistic regression and credit scorecard models) using fairly large datasets (typically ~500k records and ~1k candidate variables) to predict outcomes such as payment vs. no payment, defaulting on a loan vs. not defaulting on a loan, fraud vs. no fraud, etc. Currently, I use both SAS and R to perform feature selection using methods such as Least Absolute Shrinkage and Selector Algorithm (LASSO) regression, Random Forest, and Gradient Boosting as well as a number of clustering algorithms and principal component analysis. As such, I am constantly searching for more efficient and effective ways of building better models, which has brought me to this post.
I am very interested in learning about how Python and/or Java can make my life easier/more interesting in the realm of predictive modeling, and I am hoping that someone can provide me with the following:
1) What are some reasons to learn Python and/or Java to become a better predictive modeler?
2) Are there any references that you would recommend for me to learn more about this topic?
3) Which program would be more useful for me to learn, Python or Java (or some other programming language)?
I really appreciate any help and/or insight that you can provide! Thanks!