Artificial Intelligence MCQs with Answers
Practice important Artificial Intelligence MCQs with answers and explanations.
Chapter: 41
Q1261:
Mean Squared Error (MSE) is calculated during which phase of ANN?
A. Feature representation
B. Training
C. Validation
D. Testing
Correct Answer:
B
Chapter: 41
Q1262:
Validation data is used during training to ________.
A. Test final accuracy
B. Adjust weights
C. Avoid overfitting
D. Initialize weights
Correct Answer:
C
Chapter: 41
Q1263:
What does the radius of the neighborhood around the BMU affect?
A. Number of clusters
B. Weight adjustment of neighboring nodes
C. Input size
D. Learning rate
Correct Answer:
B
Chapter: 41
Q1264:
In the MATLAB code provided, how many features are used to represent each image?
A. 50
B. 100
C. 512
D. 1024
Correct Answer:
B
Chapter: 41
Q1265:
The MATLAB function "newff" is used to create a ________.
A. New training set
B. Neural network
C. Feature vector
D. Error metric
Correct Answer:
B
Chapter: 41
Q1266:
What training parameter defines the maximum number of iterations in the MATLAB ANN training?
A. epochs
B. goal
C. learning rate
D. momentum
Correct Answer:
A
Chapter: 41
Q1267:
Percentage accuracy on training data is calculated by comparing predicted outputs with ________.
A. Input vectors
B. Training labels
C. Target data
D. Error count
Correct Answer:
C
Chapter: 41
Q1268:
The process of grouping unlabeled data into sets based on similarity is called ________.
A. Classification
B. Regression
C. Clustering
D. Prediction
Correct Answer:
C
Chapter: 41
Q1269:
Back propagation algorithm updates weights in order to ________.
A. Maximize error
B. Minimize error
C. Randomize error
D. Ignore error
Correct Answer:
B
Chapter: 41
Q1270:
In the context of ANNs, which is NOT a typical activation function?
A. tansig
B. logsig
C. linear
D. exponential growth
Correct Answer:
D
Chapter: 41
Q1271:
Which of these is NOT part of the supervised learning process?
A. Input/output pairs
B. Finding a rule
C. Labeling data
D. Grouping unlabeled data
Correct Answer:
D
Chapter: 41
Q1272:
The term 'epoch' in neural networks means ________.
A. Number of input features
B. One complete pass through training data
C. Number of neurons
D. Number of clusters
Correct Answer:
B
Chapter: 41
Q1273:
The hypothesis space in training contains ________.
A. All possible functions to fit data
B. Only labeled data
C. Input features
D. Output classes
Correct Answer:
A
Chapter: 42
Q1274:
What is Planning in AI primarily based on?
A. Probability theory
B. Logic representation
C. Neural networks
D. Genetic algorithms
Correct Answer:
B
Chapter: 42
Q1275:
Planning in AI generates a sequence of __________ that guarantee the goal.
A. states
B. actions
C. operators
D. predicates
Correct Answer:
C
Chapter: 42
Q1276:
In Planning, a sequence of actions is commonly called a _______.
A. chain
B. flow
C. plan
D. strategy
Correct Answer:
C
Chapter: 42
Q1277:
Which of the following is NOT an improvement of Planning over classical problem solving?
A. Using predicate logic for states
B. Representing goal as predicates
C. Ignoring preconditions
D. Using backward chaining
Correct Answer:
C
Chapter: 42
Q1278:
In Planning, each state is represented as a conjunction (AND) of ________.
A. predicates that are true
B. operators
C. random variables
D. conditions
Correct Answer:
A
Chapter: 42
Q1279:
What does an action/operator in planning require to be true before it can be applied?
A. Post-conditions
B. Preconditions
C. Effects
D. Goals
Correct Answer:
B
Chapter: 42
Q1280:
What is used in Planning to avoid trying illogical operators?
A. Random choice
B. Blind search
C. Logical acting
D. Guesswork
Correct Answer:
C