Neural Networks Quiz

What is the main role of synapses in a biological neural network?

• To transmit electrical signals between neurons

• To store information permanently

• To generate new neurons

• To control the flow of blood in the brain

Which of the following best describes Hebbian learning?

• Error-based learning

• "Cells that fire together, wire together"

• Random weight adjustment

• Learning through backpropagation

The Perceptron learning rule is mainly used for which of the following?

• Nonlinear classification

• Linear classification

• Clustering

• Regression tasks

What is the key limitation of the Perceptron learning rule?

• It cannot handle large datasets

• It does not work for linearly separable problems

• It works only for linearly separable problems

• It is too computationally expensive

How are weights updated in the Perceptron learning rule?

• By increasing them randomly

• Using the gradient of the loss function

• By adding the product of the learning rate, error, and input

• Using Hebbian learning

Which of the following is true about Hebbian learning?

• It strengthens the synaptic connection when two neurons are inactive

• It weakens synaptic connections over time

• It strengthens the connection when two neurons activate simultaneously

• It adjusts weights based on the output error

In artificial neural networks, which part is functionally similar to the cell body (soma) of a biological neuron?

• Weights

• Activation function

• Bias

• Input layer

What happens when a Perceptron makes a wrong prediction during training?

• The weights are reset to zero

• The weights are updated using an error correction rule

• The input data is discarded

• The activation function is changed

What is the main difference between single-layer and multi-layer neural networks?

• The number of neurons in the output layer

• The presence of hidden layers in multi-layer networks

• The use of different activation functions

• The number of inputs to the network

Which of the following is true about a single-layer Perceptron?

• It can solve both linear and nonlinear classification problems

• It is only capable of solving linearly separable problems

• It uses backpropagation for learning

• It has multiple hidden layers