We have recently implemented the method Feature Binning and Quantile Transformation. Due to upgraded data preparation, our ML models now achieve better results.