I am getting better results when I enter as inputs both Z-score standardization and Min-Max scaling. The only worry is when it comes to data selection the algorithm is some time selecting the same feature for both Z-score and Min-Max. Is it a problem? and How to justify that scientifically ??