動画検索
関連広告
検索結果
Running the notebook
Overview of content
Relational databases as graphs
Tabular datasets as graphs
Q: Better to have more nodes or more relationships?
Q: Can we use multiple edge types?
Q: How do you build similarity-based connections?
Q: How to decide between representing a column as a node or node property?
Configuring the notebook environment
Looking at example fraud data to design the graph
Creating the nodes
Creating the edges
Instantiating the graph
Attaching labels to the nodes
Visualizing the graph
Constructing node features
Baseline (non-graph) models
Label propagation
Graph neural network
Handling featureless nodes
Q: How to figure out if test nodes are connected to train nodes?
Q: How to split into train/test
An alternative graph structure and time-directed edges
Q: On the interaction of time and graph structure
Q: Should I remove strong features to troubleshoot the model?