Transductive learning
Transductive learning infers representations or predictions for specific entities and structures observed during training rather than learning rules that generalize to unseen entities.
Optimization Algorithms: AI techniques for design, planning, and control problems
Transductive embedding is a graph embedding technique that learns representations only for entities observed during training.
- How it is used
- Learns node representations for a fixed graph when all relevant nodes are available during training.
- How deep it goes
- Defines the training restriction, scalability drawback, and static-graph fit but does not detail individual algorithms.
- Examples
- Names node2vec, DeepWalk, graph factorization, and HOPE as transductive graph-embedding methods.
- What you can do after
- Evaluate whether node embeddings fit a static graph or require inductive generalization to unseen nodes.
Graph Neural Networks in Action
A transductive embedding method directly learns representations for nodes in a fixed graph and does not naturally generalize to unseen nodes or graphs.
- How it is used
- Supplies node embeddings for downstream tasks on a known, relatively stable graph.
- How deep it goes
- Defines per-node optimization and full-graph training, then explains retraining and scaling trade-offs without deriving the optimization.
- Examples
- Uses political-books recommendations, social-network community detection, and incoming-email spam detection to illustrate suitable and unsuitable settings.
- What you can do after
- Evaluate/understand when fixed-graph embeddings fit a task and when unseen nodes favor an inductive method.
Read how Graph Neural Networks in Action explains transductive learning →
Knowledge Graphs and LLMs in Action
Transductive learning learns representations or predictions for the specific nodes and graph structure observed during training rather than a general rule for unseen nodes.
- How it is used
- Applies graph learning to classify or link known nodes when the graph is relatively static.
- How deep it goes
- Defines fixed-node optimization and contrasts it with inductive mappings, but stops short of implementation details.
- Examples
- Uses partially labeled nodes, known-node edge prediction, e-commerce recommendations, and a karate club membership example.
- What you can do after
- Evaluate whether a graph-learning problem can tolerate retraining when new nodes appear.
Read how Knowledge Graphs and LLMs in Action explains transductive learning →
Related concepts
- inductive inferencecontrasts with · Inductive inference is the process of deriving general rules, hypotheses, mappings, or predictions from observed examples or data.
- inductive embeddingalternative to · Inductive embedding is a graph representation technique that learns a function for generating embeddings of previously unseen entities or graphs.
- Node2Vecused by · Node2Vec is a graph representation-learning algorithm that learns node embeddings from biased random walks and their context relationships.
- embeddingkind of · An embedding is a numerical vector representation of an item that encodes features or relationships useful for computational tasks.
- vectorizationkind of · Vectorization is representing data or computations in vector or array form so they can be processed collectively rather than as separate scalar values.
These guides are generated from the text of Manning books by Manning's knowledge graph and link to the sections they draw on. The full books are available on liveBook and at manning.com.