You simply need different hyperparameters for the two options to give similar results.

In most problem cases, **standardization **and **normalization** might simply help.

**For clustering**, standardization may be quiet crucial in order to compare similarities between features based on certain distance measures. Principal Component Analysis (PCA), but we prefer standardization over Min-Max scaling since we are interested in the components that maximize the variance.

There is a disadvantage of normalization over standardization is that it dissipates some information in the data, especially outliers.

**For example:**

In the above image, scaling clusters is quite close together, that is unwanted in our case. It might cause algorithms such as **gradient descent** to take longer to converge on the same solution.

"Normalizing variables" is incorrect here. The correct term here is "normalizing/scaling the features".

Hope this answer helps.