Raw data often contains features with different scales and units, causing algorithms to weight larger-magnitude features disproportionately. Standardization (Z-score normalization) transforms features to zero mean and unit variance by subtracting the mean and dividing by standard deviation. Normalization (min-max scaling) rescales values to a fixed range, typically [0,1], by subtracting the minimum and dividing by range. Distance-based algorithms (KNN, SVM, K-Means) require scaling; tree-based models don't. Standardization preserves outlier information but creates values outside typical ranges; normalization bounds values but is sensitive to extreme values. Robust scaling uses median and interquartile range for outlier-resistant transformation. Feature scaling must use training…
Read the full lesson
Sign up free: one personalized lesson every day, matched to your role and goals.
Already have an account? Sign in