Quiz on Autoencoders in Deep Learning

What type of algorithm is an autoencoder in deep learning?

• Supervised

• Unsupervised

• Reinforcement

• Semi-supervised

What is the purpose of an autoencoder?

• Model evaluation

• Exact input-output mapping

• Feature detection

• Data augmentation

Which application involves identifying data points that do not resemble the rest of the dataset?

• Data Compression

• Anomaly Detection

• Data Generation

• Denoising

What is the code layer in an autoencoder?

• The hidden layer

• The output layer

• The input layer

• The latent space representation

Which hyperparameter determines the number of nodes in the code layer?

• Number of layers

• Code size

• Loss function

• Number of nodes per layer

Which type of autoencoder is used for reducing the number of active neurons in the coding layer?

• Sparse Autoencoder

• Regularized Autoencoder

• Denoising Autoencoder

• Stacked Autoencoder

What is the purpose of adding noise to the input in a denoising autoencoder?

• To increase overfitting

• To decrease underfitting

• To train it to recover the original input

• To improve data quality

Which type of regularization is used in a sparse autoencoder to add sparsity loss?

• L2 Regularization

• L1 Regularization

• Dropout

• Kullback-Leibler Divergence

What is the main component of a vanilla autoencoder?

• Encoder

• Code

• Decoder

• Loss function

What is the total loss in a Variational Autoencoder composed of?

• Regularization loss only

• Reconstruction loss and Kullback-Leibler divergence

• Kullback-Leibler divergence only

• Reconstruction loss only