Data Mining – Normalization Problem
Question:
Suppose that the data for analysis includes the attribute cost price and the values for the data
tuples are:
100, 150, 140, 115, 190, 120, 130, 125, 135, 145, 140, 150, 165, 160, 170
(i) Use min–max normalization to transform the value of 145 for cost price onto the range [0,1].
(ii) Use Z-score normalization to transform the value 145 for cost price where the standard
deviation of cost price is 120.
Solution:
(i) Min–Max Normalization
Formula:
v' = (v − min(A)) / (max(A) − min(A))
min(A) = 100, max(A) = 190, v = 145
v' = (145 − 100) / (190 − 100) = 45 / 90 = 0.5
Normalized value = 0.5
(ii) Z-Score Normalization
Formula:
v' = (v − µ) / σ
Mean (µ) = 142.33, Standard deviation (σ) = 120
v' = (145 − 142.33) / 120 ≈ 0.022
Z-score normalized value ≈ 0.022