Deep Learning Quiz - chap5

What is the main purpose of using graph data in deep learning?

• Representing structured data

• Analyzing numerical data

• Creating visualizations

• Representing unstructured and complex data

Which industry does not commonly use graph data in deep learning applications?

• Agriculture

• Healthcare

• Social Networks & Recommender Systems

• Chemistry & Pharmacy

What is the purpose of Laplacian matrix in representing graphs?

• Counts number of connections to each node

• Shows the smoothness of the graph

• Represents node features

• Aggregates neighborhood features

Which of the following is not a popular approach in graph neural networks?

• Node Embedding Networks (NEN)

• Message Passing Neural Networks (MPNN)

• Graph Convolution Networks (GCN)

• Graph Attention networks (GAT)

What is the purpose of mean-pooling in graph neural networks?

• Encodes node features to embeddings

• Normalizes the vector to prevent feature explosion

• Aggregates neighborhood features

• Updates node features with information from distant neighborhoods

Which method is used to get signals from more distant neighborhoods in the graph?

• Mean-Pooling

• Message Passing Neural Networks

• Stacking Layers

• Symmetric Normalization

How is training done in graph neural networks?

• Semi-supervised learning

• Reinforcement learning

• Unsupervised learning

• Supervised learning

What is the main focus of message passing neural networks in graph data?

• Node-wise mechanisms

• Graph-level features

• Global features

• Edge-wise mechanisms

What does the adjacency matrix in graph neural networks represent?

• The neighborhood of each node

• The latent space of the graph

• The degree of each node

• The smoothness of the graph

Which task involves predicting the time to travel from one location to another in graph data?

• Graph-level task

• Edge-level task: Classification

• Edge-level task: Link Prediction

• Node-level task: Regression