Which technique squashes data into a fixed range between zero and one?

Which technique squashes data into a fixed range between zero and one?

• A. Normalization (Min-Max)
• B. Standardization
• C. Data Transformation
• D. Principal Components Analysis (PCA)

Answer: (A)

Min-max normalization is the technique that maps each feature to the 0–1 interval. It rescales values using the formula (x − min(feature)) / (max(feature) − min(feature)). This linear transformation ensures the smallest value becomes 0 and the largest becomes 1, with all other values positioned proportionally between them. This fixed range is especially helpful for models that rely on distance or on inputs being on a similar scale, such as neural networks and k-nearest neighbors, to prevent any single feature from dominating. Standardization centers data around zero and scales to unit variance, but it does not bound values to [0,1], so it isn’t the technique described. PCA focuses on reducing dimensionality by projecting onto principal components and doesn’t inherently bound data to a fixed range. Data transformation is a broad term and doesn’t specify the fixed-range outcome. Keep in mind that outliers can stretch the min and max and affect the scaling, in which case alternative scaling methods may be preferred.

Min-max normalization is the technique that maps each feature to the 0–1 interval. It rescales values using the formula (x − min(feature)) / (max(feature) − min(feature)). This linear transformation ensures the smallest value becomes 0 and the largest becomes 1, with all other values positioned proportionally between them. This fixed range is especially helpful for models that rely on distance or on inputs being on a similar scale, such as neural networks and k-nearest neighbors, to prevent any single feature from dominating.

Standardization centers data around zero and scales to unit variance, but it does not bound values to [0,1], so it isn’t the technique described. PCA focuses on reducing dimensionality by projecting onto principal components and doesn’t inherently bound data to a fixed range. Data transformation is a broad term and doesn’t specify the fixed-range outcome. Keep in mind that outliers can stretch the min and max and affect the scaling, in which case alternative scaling methods may be preferred.