AI0-001 AI Models and Data Engineering Practice Question

Answer analysis

Option-by-option breakdown

For each option: why learners choose it and why it is or isn't the right answer here.

• ✓

  Logarithmic transformation

  Why this is correct

  A logarithmic transformation compresses the right tail and can make a right-skewed distribution more symmetric. It is effective when data spans several orders of magnitude and contains positive values. This helps linear models meet assumptions of normality and reduces the impact of outliers, improving model performance.

• ✗

  Standardization (z-score normalization)

  Why it's wrong here

  Standardization rescales data to have zero mean and unit variance but does not change the shape of the distribution. It will not reduce skewness or make the distribution more symmetric. It is useful for algorithms sensitive to scale but does not address the skewness issue described.

• ✗

  Square root transformation

  Why it's wrong here

  A square root transformation is milder than logarithmic and can reduce skewness, but it is less effective for highly skewed data with a long right tail. It is better for moderate skewness or count data. For a highly skewed distribution, logarithmic transformation is more appropriate to achieve symmetry.

• ✗

  Min-max normalization

  Why it's wrong here

  Min-max normalization scales values to a fixed range, typically [0,1], but preserves the original distribution shape. It does not reduce skewness or make the distribution more symmetric. Like standardization, it is a scaling technique, not a transformation for skewness correction.