nerdexam
Amazon

MLS-C01 · Question #321

A data scientist is building a linear regression model. The scientist inspects the dataset and notices that the mode of the distribution is lower than the median, and the median is lower than the…

The correct answer is B. Logarithmic transformation. A distribution where mode < median < mean indicates a right-skewed distribution, and a logarithmic transformation is commonly used to normalize such data for linear regression models.

Modeling

Question

A data scientist is building a linear regression model. The scientist inspects the dataset and notices that the mode of the distribution is lower than the median, and the median is lower than the mean. Which data transformation will give the data scientist the ability to apply a linear regression model?

Options

  • AExponential transformation
  • BLogarithmic transformation
  • CPolynomial transformation
  • DSinusoidal transformation

How the community answered

(54 responses)
  • A
    4% (2)
  • B
    70% (38)
  • C
    17% (9)
  • D
    9% (5)

Why each option

A distribution where mode < median < mean indicates a right-skewed distribution, and a logarithmic transformation is commonly used to normalize such data for linear regression models.

AExponential transformation

Exponential transformation would increase the skewness of an already right-skewed distribution, moving it further away from a normal distribution suitable for linear regression.

BLogarithmic transformationCorrect

When mode < median < mean, the data distribution is positively or right-skewed. A logarithmic transformation (e.g., natural log or log base 10) is a standard technique to reduce the skewness in right-skewed data, making its distribution more symmetric and closer to normal, which is often a desirable assumption for linear regression models.

CPolynomial transformation

Polynomial transformation primarily helps capture non-linear relationships between variables but does not inherently correct for skewness in the distribution of a single variable to make it more Gaussian.

DSinusoidal transformation

Sinusoidal transformation is used for cyclical or periodic data, which is not indicated by the described skewness of the distribution.

Concept tested: Data transformation for skewed distributions

Source: https://aws.amazon.com/blogs/machine-learning/optimizing-machine-learning-predictions-with-data-transformations-part-1-power-transforms/

Topics

#Data Transformation#Skewness#Linear Regression Preprocessing#Statistical Distributions

Community Discussion

No community discussion yet for this question.

Full MLS-C01 Practice