Performance of the accuracy score as a function of different hyperparameters for the training sample (red line) and testing sample (blue line). In the top left panel we show the accuracy for different scalers: standard (1), MinMax (2), MaxAbs (3), robust (4), power transformer (5), quantile transformer (6), normalizer L1 (7), normalizer L2 (8), and normalizer-max (9). Marked in red (7) the one adopted in the present work.