Question 1

When preparing input data for a predictive model, which of the following issues is likely to cause the most significant problems in the performance of the model?

Options :

A :

The input data contains several variables with a small proportion of missing values that can be easily imputed.

B :

The input data contains a few high-cardinality categorical variables that require encoding.

C :

There is a considerable amount of multicollinearity present among the predictor variables.

D :

The input data has been standardized using z-scores.

Answer: C

Question 2

A binary classifier is used to predict a rare event. Its performance is summarized in the following confusion matrix: | | Predicted Negative | Predicted Positive | |-||| | Actual Negative | 9750 | 250 | | Actual Positive | 25 | 75 | Given the model's performance, what is the False Positive Rate (FPR) of the classifier?

Options :

A :

0.0256

B :

0.03

C :

0.25

D :

0.0025

Answer: A

Question 3

A data scientist is using the SAS system to preprocess data for predictive modeling. They need to convert a character variable named `char_date` in the format 'yyyy-mm-dd' to a SAS date variable. Which INPUT statement below correctly performs this conversion?

Options :

A :

date_var = INPUT(char_date, yymmdd10.);

B :

date_var = INPUT(char_date, date9.);

C :

date_var = INPUT(char_date, yymmddd10.);

D :

date_var = INPUT(char_date, ddmmyy10.);

Answer: A

Question 4

You are developing a predictive model using SAS and have completed the model training phase. You are now preparing to deploy the model for scoring a new dataset. Which mode should you operate in to score the new dataset and why?

Options :

A :

Training mode, because the model needs to be retrained on the new dataset before it can be scored.

B :

Training mode with cross-validation, to ensure that the model generalizes well to the new dataset.

C :

Output mode, as the model is already trained and the goal now is to use the model to predict new outcomes.

D : Output mode, but with added regularization parameters to avoid overfitting on the new dataset.

Answer: C

Question 5

You are performing predictive modeling using a random forest technique in R through SAS Enterprise Miner. You want to examine variable importance to interpret the model. Which of the following commands within the R code can provide you with the variable importance measure typically associated with a random forest model?

Options :

A :

randomForest::importance(model)

B :

rpart::varImp(model)

C :

caret::varImp(model, scale = FALSE)

D :

varImpPlot(model)

Answer: A

Confidently Practice Online with Free A00-225 Exam Cram

Buying Options: