G12 A - Term 2 Quiz Revision

What do we call a model that cannot model the training data or generalise to new data?

• Well- trained model

• Validation dataset

• Underfit model

• Overfit model

Illustrate a multi-layer perceptron neural network.

• A neural network with only one layer

• A neural network with multiple layers

• A neural network with no layers

• A neural network with random layers

What do we call a model that can produce high accuracy on both testing and validation data?

• Underfit model

• Overfit model

• Well- trained model

Which type of neural networks works on the principle of saving the output of a layer and feeding it back to the input to predict the outcome of the layer?

• Deep Feedforward Neural Network

• Recurrent Neural Network

• Feedforward Neural Network

• Single Layer Perceptron

Which of the following refers to a sample of data used to fit the model?

• Training dataset

• Test dataset

• Sample dataset

Which of the following problems cannot be directly solved by neural networks?

• Classification problems

• Clustering problems

• Hardware problems

• Regression problems

Examine different types of neural networks.

• Feedforward, Backpropagation, Perceptron, Support Vector Machine

• Convolutional, Recurrent, Deep Belief, Kohonen

• Regression, Classification, Clustering, Dimensionality Reduction

• Linear, Polynomial, Exponential, Logarithmic

Which of the following is NOT an application of neural networks?

• Image processing

• Machine translation

• Speech Recognition

• Machine engines

Demonstrate basic concepts and methods in the field of computer vision.

• Segmentation, Feature extraction, Object detection

• Algebra, Calculus, Geometry

• Literature analysis, Historical research, Linguistics

• Music theory, Art appreciation, Dance techniques

What do we call a layer that is deeply connected to the layer before it?

• Input layer

• Output layer

• Dense layer