I am working on my Master's thesis statement, in which I am using a bank's transaction dataset. I am so confused whether I should use robust scaler or standard scaler so as to scale transaction amounts. The variance of features would be so different in each scale method.