Artificial Intelligence MCQs with Answers
Practice important Artificial Intelligence MCQs with answers and explanations.
Chapter: 40
Q1241:
Fill in the blank: In a ______ neural network, neurons are connected in a loop.
A. Feed-forward
B. Recurrent
C. Single layer
D. Multilayer
Correct Answer:
B
Chapter: 40
Q1242:
Fill in the blank: A ______ neural network has only one input layer and one output layer.
A. Single layer
B. Multilayer
C. Feed-forward
D. Recurrent
Correct Answer:
A
Chapter: 40
Q1243:
Fill in the blank: The term ______ describes the strength of connection between neurons.
A. Synaptic weight
B. Neuron bias
C. Learning rate
D. Activation function
Correct Answer:
A
Chapter: 40
Q1244:
Which of the following is a drawback of ANNs?
A. Data dependency
B. Excellent for classification
C. Handles noisy data
D. Parallel architecture
Correct Answer:
A
Chapter: 40
Q1245:
Which neural network type allows signal to flow in one direction only?
A. Feed-forward
B. Recurrent
C. Partially connected
D. Fully connected
Correct Answer:
A
Chapter: 40
Q1246:
What is NOT a component of a multilayer network?
A. Input layer
B. Hidden layer
C. Decision layer
D. Output layer
Correct Answer:
C
Chapter: 41
Q1247:
What type of learning involves given a set of example input/output pairs to find a rule predicting outputs for new inputs?
A. Supervised learning
B. Unsupervised learning
C. Reinforcement learning
D. Clustering
Correct Answer:
A
Chapter: 41
Q1248:
In unsupervised learning, data is grouped into sets called ________.
A. Clusters
B. Classes
C. Rules
D. Features
Correct Answer:
A
Chapter: 41
Q1249:
Where is knowledge primarily encoded in Artificial Neural Networks (ANNs)?
A. In the input data
B. In the network architecture
C. In the weights of neurons
D. In the output layer
Correct Answer:
C
Chapter: 41
Q1250:
Which phase of ANN design involves determining the number of features based on the problem inputs?
A. Training
B. Feature representation
C. Validation
D. Activation function
Correct Answer:
B
Chapter: 41
Q1251:
What is the goal of training in neural networks?
A. To find the best hypothesis fitting most training examples
B. To memorize training data
C. To generate random outputs
D. To increase network size
Correct Answer:
A
Chapter: 41
Q1252:
Which function is commonly used to measure total error during neural network training?
A. Mean Absolute Error
B. Cross Entropy
C. Mean Squared Error
D. Root Mean Squared Error
Correct Answer:
C
Chapter: 41
Q1253:
During training, what method helps avoid overfitting by dividing training data into k sets?
A. Validation
B. Clustering
C. Randomization
D. Weight updation
Correct Answer:
A
Chapter: 41
Q1254:
What is a common stopping criterion in ANN training?
A. Minimum Mean Squared Error threshold
B. Maximum number of epochs
C. Both A and B
D. None of these
Correct Answer:
C
Chapter: 41
Q1255:
A network that correctly generalizes input-output relationships for unseen data is said to have good ________.
A. Validation
B. Training
C. Generalization
D. Clustering
Correct Answer:
C
Chapter: 41
Q1256:
Which algorithm randomizes weights to small positive and negative values initially?
A. Gradient Descent
B. Back Propagation
C. K-means Clustering
D. Decision Tree
Correct Answer:
B
Chapter: 41
Q1257:
In backpropagation, the error is the difference between the desired output and the ________.
A. Input vector
B. Network output
C. Weight vector
D. Learning rate
Correct Answer:
B
Chapter: 41
Q1258:
In unsupervised learning, what is the "Best Matching Unit" (BMU)?
A. The node closest to the input vector
B. The node with the highest weight
C. The node with the lowest error
D. The node with the smallest radius
Correct Answer:
A
Chapter: 41
Q1259:
Which type of neural network is used for clustering and function approximation?
A. Feedforward
B. Convolutional
C. Self-organizing
D. Recurrent
Correct Answer:
C
Chapter: 41
Q1260:
Fill in the blank: The process of adjusting weights in a neural network during training is called ________.
A. Activation
B. Weight updation
C. Validation
D. Initialization
Correct Answer:
B