Text as Data (H) - Lecture 07 Quiz
1.
Sports - 28/55
Not sports - 27/1000
Not sports - 108/10000
Sports - 4/100
2. You are asked to build a supervised learning technique to identify all of the objects present in a picture. Is this:
regression
multi-label multi-class classification
single-label multi-class classification
binary classification
3. Which classifier is infinitely flexible, able to fit to any features?
None of the above
Logistic Regression
Naive bayes
Decision Tree
4. In picking the next decision point, a decision trees picks the feature that
best discriminates between the classes
balances the classes across the decision
5. Why is smoothing used in Naive Bayes?
So the non-occurrence of a term does not result in 0 probability
To reduce floating point operations
So that our numbers round easily to nice fractions
Submit Quiz