Machine Learning Mid  I, September  2014
1.Machine learning is inherently a multidisciplinary field.

Inter disciplinary

Multidisciplinary

single

None

Answer: B
2.A computer program is said to learn from __________E with respect to some class of tasks T and performance measure P, if its performance at tasks in T, as measured by P, improves with experience E.

Training

Experience

Database

Algorithm

Answer: A
3.__________methods have been used to train computercontrolled vehicles to steer correctly when driving on a variety of road types.

Machine Learning

Data Mining

Neural networks

Robotics

Answer: A
4.Any hypothesis found to approximate the target function well over a sufficiently large set of training examples will also approximate the target function well over other unobserved examples

Hypothesis

Inductive Hypothesis

Learning

Concept Learning

Answer: B
5.The _________algorithm computes the version space containing all hypotheses from H that are consistent with an observed sequence of training examples.

Inductive Hypothesis

Artificial Neural Network

Candidate Elimination

none

Answer: C
6.The _________, denoted VSHVD, with respect to hypothesis space Hand training examples D, is the subset of hypotheses from H consistent with the training examples in D.

Space

Vertical space

version space

version spain

Answer: C
7.Quinlan and Rivest (1989) describe experiments applying the MDL principle to choose the _______ for a decision tree.

Best size

big size

Small size

over fit

Answer: A
8.Minimum Description Principle is a version of _________that can be interpreted within a Bayesian Network.

ID3

Selection measure

occam’s razor

PAC

Answer: C
9.A perceptron takes a vector of realvalued inputs, calculates a linear combination of these inputs, then outputs ____________

1 or 1

0 or 1

1 or 0

none

Answer: A
10.If the training examples are not linearly separable, the delta rule converges toward a approximation to the target concept.

Over fit

under fit

doesn’t fit

best fit

Answer: D
11.Theoretical results have been developed that characterize the fundamental relationship among the number of __________________ examples observed.
Answer: Training
12.The __________of L is any minimal set of assertions B such that for any target concept c and corresponding training examples Dc.
Answer: Inductive Bias
13.To apply MDL principle in practice we must choose _______________appropriate for the given learning task.
Answer: Specific encodings or representations
14._____________________Prefer the simplest hypothesis that fits the data.
Answer: Occam’s razor
15._____________is a significant practical difficulty for decision tree learning and many other learning methods.
Answer: Overfitting
16.One successful method for finding high accuracy hypotheses is a technique called___________.
Answer: Postpruning
17.__________learning methods provide a robust approach to approximating realvalued, discretevalued, and vectorvalued target functions.
Answer: Neural network
18.In learning to play checkers, the system might learn from ________training examples consisting of individual checkers board states and the correct move for each.
Answer: Direct
19.learning algorithms to acquire only some approximation to the target function, and for this reason the process of learning the target function is often called __________________.
Answer: Function approximation
20.PAC Acronym ___________________________________.
Answer: Probably Approximately Correct