MS-DS Master of Data science MS-DS Master of Data science Data Wrangling and Preprocessing Questions and Answers 2 — Questions and Answers
Question 1: Which technique is most appropriate for handling multicollinearity among predictor variables during preprocessing?
- Variance Inflation Factor analysis followed by feature removal (Correct answer)
- Min-max normalization
- Label encoding
- Forward fill imputation
Correct answer: Variance Inflation Factor analysis followed by feature removal
VIF analysis quantifies multicollinearity and guides removal of highly correlated predictors to stabilize model coefficients.
Question 2: When working with a dataset that has 40% missing values in a single feature, which approach is generally preferred over simple deletion?
- Multiple imputation (Correct answer)
- Listwise deletion
- Dropping the entire feature immediately
- Replacing all missing values with zero
Correct answer: Multiple imputation
Multiple imputation generates several plausible datasets to account for uncertainty in missing data, outperforming simple deletion when missingness is substantial.
Question 3: What is the primary purpose of applying a Box-Cox transformation during data preprocessing?
- To make the data distribution more closely approximate a normal distribution (Correct answer)
- To remove duplicate records from the dataset
- To encode categorical variables as integers
- To split the dataset into training and testing sets
Correct answer: To make the data distribution more closely approximate a normal distribution
Box-Cox transformation applies a power transformation to stabilize variance and make data more normally distributed for parametric modeling.
Question 4: In a data wrangling pipeline, what does the term 'tidy data' specifically require?
- Each variable forms a column, each observation forms a row, and each type of observational unit forms a table (Correct answer)
- All missing values are removed before analysis
- Column names are standardized to snake_case
- Data is sorted in ascending order by the primary key
Correct answer: Each variable forms a column, each observation forms a row, and each type of observational unit forms a table
Tidy data, as defined by Hadley Wickham, requires variables in columns, observations in rows, and each observational unit in its own table.
Question 5: Which method is best suited for detecting outliers in a multivariate dataset?
- Mahalanobis distance (Correct answer)
- Z-score on individual columns
- Replacing values beyond the IQR
- Histogram visual inspection
Correct answer: Mahalanobis distance
Mahalanobis distance accounts for correlations between variables, making it effective for identifying multivariate outliers that univariate methods miss.
Question 6: What is the key advantage of using robust scaling (median and IQR) over standard scaling (mean and standard deviation)?
- Robust scaling is less affected by outliers in the data (Correct answer)
- Robust scaling always produces values between 0 and 1
- Robust scaling converts categorical features to numerical
- Robust scaling eliminates the need for missing value imputation
Correct answer: Robust scaling is less affected by outliers in the data
Robust scaling uses the median and interquartile range, which are resistant to extreme values, unlike the mean and standard deviation used in standard scaling.
