A machine learning algorithm may benefit from data that is "standardized". If one column in a dataframe has a completely different scale than the rest, it may cause an overfit. To mitigate this you may consider scaling your dataset. This video will give a full introduction on what this does to your pipeline.