Question 2 To Evaluate classifiers, we use methods that are different from methods used to evaluate clustering.

a. True

b. False

Question 3 If recall equals 1, and precision equals 0.5, F1 would be

a. 0.75

b. 1/1.5

c. 1.5

d. None are correct

Question 4 Which of the following is not correct

a. Accuracy is not commonly used when labels are not equally distributed in the dataset

b. The data in Data warehouse is not generated fast

c. In Artifical neural networks, weights are updated when the predicted label equals the actual label

d. All are correct

Question 5 Instead of using student grade, we want to use a range such as 50-60, 60-70, ... This is called a. Attribute Construction

b. Data Aggregation

c. Discretization

d. Removing Redundancy

Question 6 We want to perform scaling of the field Tempreature. If values were origionally from 0 to 40 and we want to convert them to 0 to 1. The value 10 will be

a. 0.25

b. 0.3

c. 0.4

d. None of the above

Question 7

Some mobile applications allow you to take a photo of any important building and the application will recognize the building name based on historical training data. This is called

a. Classification

b. Association Rule Discovery

c. Clustering

d. Anomaly Detection

e. Not Data Mining

Question 8

If Mean TF.IDF of word Win is 2. The total number of documents is 3. TF.IDF of win in doc1 = 4 and in doc2 = 0. If the word win appears in the three documents, find frequency of win in doc3.

a. 2

b. 4

c. We can not know

d. None of the above