Neural Networks & Deep Learning Assessment

Test concepts of neural networks, deep learning architectures, training techniques, optimization methods, and real-world applications using modern frameworks.

Verified Assessment

Assessment Details

Questions 20 Questions

Passing Score 60% Correct

Time Limit 30 Minutes

Format Multiple Choice

Rules

Runs in Full Screen mode

Do not switch tabs

Do not minimize window

Warning: Violations will auto-submit the quiz immediately.

Clicking start will enter full-screen mode.

Neural Networks & Deep Learning Assessment

Assessment Details

Rules

A convolution layer applies a 3×3 kernel to a 5×5 input image with stride 1 and no padding. What is the output feature map size?

Which loss function is used for multi-class classification?

What does backpropagation compute?

What is the role of batch normalization?

What does transfer learning allow?

Which activation function is most commonly used in hidden layers?

Which framework is widely used for dynamic computation graphs?

What problem do residual connections solve?

What is the main purpose of an activation function?

Why do deep networks suffer from vanishing gradients?

Which technique helps reduce overfitting?

A deep learning model shows low training loss but high validation loss. What does this indicate?

What is the output of the ReLU activation function for input -3?

If the learning rate is set too high, what is the most likely training behavior?

What is the main purpose of an autoencoder?

Which architecture is best suited for image data?

What does overfitting mean in deep learning?

A neuron has inputs [1, 2], weights [0.5, 1.0], and bias 0. What is the output before activation?

A binary classification model predicts a probability of 0.9 for class 1. Using a threshold of 0.5, what is the predicted class?

Which optimizer adapts learning rates per parameter?

Security Violation

Assessment Terminated

Quiz Not Passed

Sorry, You Didn't Pass