Question 1

You are working with a logistic regression model for a binary classification problem. You notice that the training error is significantly lower than the validation error. What could you do to improve the model's performance on the validation set?

Options :

A : Increase the model-s learning rate.

B : Add more features to the model.

C : Remove all features and retrain the model.

D : Regularize the model by introducing L1 or L2 penalties.

Answer: D

Question 2

You are analyzing the relationship between hours of study and exam scores among students. After calculating, you find a correlation coefficient (r) of 0.9. Which of the following is the best interpretation?

Options :

A : Hours of study has a strong, positive correlation with exam scores

B : Hours of study has a moderate, negative effect on exam scores

C : Exam scores have no relationship with hours of study

D : Hours of study causes high exam scores

Answer: A

Question 3

You have a dataset of a retail company's sales transactions, which includes details like customer ID, transaction amount, transaction date, and store ID. You're tasked with summarizing monthly revenue for each store. What would be the most appropriate data aggregation technique?

Options :

A : Group the data by store ID and sum the transaction amounts daily, then take the average to get the monthly revenue.

B : Group the data by customer ID and store ID, then sum the transaction amounts and divide by the number of unique customers for each month.

C : Group the data by transaction date and calculate the daily revenue, then aggregate this to the monthly level without considering store ID.

D : Group the data by store ID and sum the transaction amounts for each month.

Answer: D

Question 4

You have a Pandas DataFrame that contains missing values. Which of the following methods will replace all NaNs with zeros?

Options :

A : df.fillna(value=0)

B : df.replace(to_replace=np.nan, value=0)

C : df['column_name'] = 0

D : df.dropna(subset=['column_name'])

Answer: A

Question 5

What is the most appropriate technique to handle missing values for a categorical column in a dataset before fitting it to a model?

Options :

A : Remove Rows with Missing Values

B : Mean Imputation

C : Linear Regression Imputation

D : Mode Imputation

Answer: D

Confidently Practice Online with Free PCED Exam Cram

Buying Options: