George Jenks developed this optimization system. The goal is forming groups that are internally homogeneous while assuring heterogeneity among classes. This has proven to be a very useful method, next to natural breaks - but requires computing power to perform. A statistical approach based on “Min & Max” of data variance. Data variance – how much data values vary in magnitude among each other. Start with a single class: range (a single class) = max data value – min data value. Introduce another group whereby. Minimize within group variance (member data values closer in value). Maximize between group variance (difference in group averages as great as possible).
Sources from wikipedia
Tiada ulasan:
Catat Ulasan