EDA Methods DY0-001 Practice Questions (Page 2)

Questions 6–10

1. 6expert · hard

   A data scientist is analyzing a manufacturing dataset to predict machine failure. The dataset has 20 numeric sensor readings. Initial univariate analysis shows no single sensor clearly separates failures from normal operation. The scientist suspects that failures occur only when specific combinations of sensors exceed thresholds simultaneously. Which EDA technique is most effective for revealing these multivariate patterns?

   • AA correlation heatmap of all 20 sensors
   • BDimensionality reduction with PCA, then a scatter plot of the first two principal components colored by failure status
   • CBox plots of each sensor grouped by failure status
   • DA scatter plot matrix of the first 5 sensors colored by failure status
   Reveal Solution Discussion Select an answer first
2. 7expert · hard

   A data scientist is analyzing a dataset with 5,000 samples and 30 features. They create a scatter plot of two features, Feature_A and Feature_B, colored by the target class. The plot shows two distinct, well-separated clusters, but the clusters are diagonal (correlated). The scientist then computes the Pearson correlation between Feature_A and Feature_B and finds it is 0.9. What is the most important implication for feature selection?

   • AThe high correlation indicates the target classes are imbalanced
   • BThe high correlation suggests multicollinearity; one of the two features may be redundant for a linear model, but the joint pattern is valuable for non-linear models
   • CThe high correlation proves the features are identical and one is a typo
   • DThe high correlation means both features should be removed immediately to avoid data leakage
   Reveal Solution Discussion Select an answer first
4. 8application · medium

   A data analyst is examining a customer churn dataset with 10,000 records. The target variable 'churned' is binary (0/1), and the analyst needs to quickly assess whether the dataset is balanced and whether any single feature obviously separates churned from non-churned customers. Which combination of EDA techniques should the analyst use first?

   • AA scatter plot matrix of all numeric features colored by churn status
   • BHistograms of every numeric feature without grouping
   • CA correlation heatmap of all numeric features
   • DA bar chart of the 'churned' variable and box plots of each numeric feature grouped by churn status
   Reveal Solution Discussion Select an answer first
5. 9foundation · easy

   Which technique is commonly used to identify underlying patterns or groupings in a dataset with many variables?

   • AHistogram analysis
   • BMean and median calculation
   • CFrequency table
   • DPrincipal component analysis (PCA)
   Reveal Solution Discussion Select an answer first
6. 10foundation · easy

   A data scientist wants to explore how advertising spend, product price, and seasonality together influence sales. Which type of analysis is most appropriate?

   • ABivariate analysis
   • BUnivariate analysis
   • CDescriptive statistics only
   • DMultivariate analysis
   Reveal Solution Discussion Select an answer first