How do you use the Min-Max formula to scale data into the range from 0 to 1?

Conditions

• The dataset has clear min and max values.
• max is not equal to min

Reasoning, step by step

1. Identify the original value $X$ you want to normalize.
2. Determine the minimum value $\min$ and maximum value $\max$ of the feature column.
3. Substitute these values into the formula $X' = \frac{X - \min}{\max - \min}$.
4. Calculate the result to get the normalized value $X'$, which will fall within the interval [0, 1].

Example

For a house area of 1200, with a column minimum of 1000 and maximum of 1800, the calculation is $\frac{1200 - 1000}{1800 - 1000} = \frac{200}{800} = 0.25$.

Common misconceptions

• Assuming that one can directly train machine learning models using raw large-valued features without worrying about the impact of numerical scale.

Watch the explanation

Explore next

• Why do we need to normalize or scale data in machine learning?
• When should the Z-score technique be used for data normalization?