MLA-C01 Question 55
Single answerYou are developing a machine learning model to predict house prices using a dataset that includes features such as square footage, number of bedrooms, and property age. During the exploratory data analysis, you notice that the 'property age' feature has a highly skewed distribution. Which feature engineering technique should you apply to address this issue before training your model?
- A
Log transformation
- B
One-hot encoding
- C
Normalization
- D
Feature splitting
Show answer and explanation
Correct answer: A
Explanation
When dealing with skewed numerical data, applying a log transformation is a common technique to reduce skewness and make the data distribution closer to normal. This helps improve the performance of many machine learning models that assume normally distributed data. Other techniques mentioned, such as normalization or one-hot encoding, are not suitable for addressing skewness in numerical features.
- A. Correct.
Log transformation is effective for reducing skewness in data, especially when the distribution is highly skewed, as it compresses large values and expands smaller values to create a more normal distribution.
- B. Incorrect.
One-hot encoding is used for categorical variables to convert them into numerical format. It is not applicable to numerical features with skewed distributions.
- C. Incorrect.
Normalization scales numerical feature values to a specific range (e.g., 0 to 1) but does not address skewness in the data distribution.
- D. Incorrect.
Feature splitting involves dividing a single feature into multiple features but is not a method for addressing skewed distributions.