c. A telecom company has developed a classifier for detecting whether a cell-phone network base station...

  

Transcribed Image Text:

c. A telecom company has developed a classifier for detecting whether a cell-phone network base station is faulty or not. We evaluate the classifier on a test set. Here is the confusion matrix. (In the table, F means faulty and NF not faulty.) Truth i. Compute the accuracy of the classifier. F . Predicted F NF 10 5 NF 20 965 ii. What is the accuracy of a majority-class baseline? (The class not faulty is the most common in the training set.) iii. Would you say that the classifier is more useful than the majority-class baseline? Explain why or why not. c. A telecom company has developed a classifier for detecting whether a cell-phone network base station is faulty or not. We evaluate the classifier on a test set. Here is the confusion matrix. (In the table, F means faulty and NF not faulty.) Truth i. Compute the accuracy of the classifier. F . Predicted F NF 10 5 NF 20 965 ii. What is the accuracy of a majority-class baseline? (The class not faulty is the most common in the training set.) iii. Would you say that the classifier is more useful than the majority-class baseline? Explain why or why not.