ML CS7641 MIDTERM UPDATED ACTUAL QUESTIONS AND CORRECT ANSWERS

ML CS7641 MIDTERM UPDATED ACTUAL QUESTIONS
AND CORRECT ANSWERS

Question:
1. Supervised Learning
Answer:
school of machine learning that relies on human input to train a model

Question:
2. Basal assumption of supervised learning
Answer:
there exists some well-behaved, consistent function behind data we're seeing

Question:
3. Classification
Answer:
mapping complex inputs to labels/classes/discrete values

Question:
4. Regression
Answer:
mapping complex inputs to any numeric values

Question:
5. Source of data errors
Answer:
hardware, malicious intent, human element, unmodeled influences

Question:
6. Graph of Fit
Answer:




Question:
7. Where is a good fit of data

,Answer:
where the error across both training data and cross-validation data are relatively similar.

Question:
8. Cross validation
Answer:
a method used for reducing overfitting

Question:
9. Instances
Answer:
representing the input data from which the overall model will "learn"

Question:
10. Concept
Answer:
Abstract idea that represents data

Question:
11. Candidate
Answer:
potential target concept

Question:
12. Testing set
Answer:
instances that our candidate concept has not yet seen in order to evaluate how close it is to the ideal target
concept

Question:
13. Decision Trees
Answer:
Map various choices to diverging paths that end with some decision

Question:
14. Order in which features are best applied to decision trees
Answer:
correlated with its ability to reduce our VC space

Question:
15. ID3 Algorithm
Answer:
- A< -best attribute
- Assign A as decision attribute
- for each option in A, create branch n
- lump training examples to respective branches
-if perfectly classified: stop, else: repeat

, Question:
16. Information Gain Equation
Answer:




Question:
17. Information Gain
Answer:
How much an attribute can reduce overall entropy

Question:
18. Entropy Equation
Answer:




Question:
19. Entropy
Answer:
Measure of how much information an attribute gives about a system

Question:
20. "Best" Attribute
Answer:
One with maximum information gain

Question:
21. Restriction Bias
Answer:
automatically occurs when we decide our hypothesis set, H.

Question:
22. Preference Bias
Answer:
what sort of hypotheses from our hypothesis set, h ■ H

Question:
23. Preference Bias of ID3