Developed a custom code in 'R' which uses 'multi-variate' or 'oblique' splits in each of the Decision Tree instance of a Random Forest structure. The optimal split is a 'linear combination' of two or more features selected at random in a single decision tree instance (oblique) and gives better space & time complexity than the traditional 'univariate' split. Leveraged the standard implementation of the 'randomForest' package to make the oblique splits. The output (error rate) is comparable to that of the custom R packages - 'oblique.RF', 'oblique.tree', 'PPForest (Project Pursuit Forest)'. Technologies (Programming Languages) - Python R