omg i was so confused about this too! what does normalize data mean? it’s like when you’re baking and you convert cups to grams so everything’s in the same unit.
you NEED it for ML because some algorithms (like k-means or gradient descent) are super sensitive to scale. if you skip it, your model might train slower or just... suck.
min-max is simplest (just squish everything into 0-1), but z-scores are better if you have outliers. pandas + sklearn make it a breeze.
you NEED it for ML because some algorithms (like k-means or gradient descent) are super sensitive to scale. if you skip it, your model might train slower or just... suck.
min-max is simplest (just squish everything into 0-1), but z-scores are better if you have outliers. pandas + sklearn make it a breeze.