Inductive inference
Inductive inference is the process of deriving general rules, hypotheses, mappings, or predictions from observed examples or data.
Knowledge Graphs and LLMs in Action
Inductive reasoning is the derivation of general patterns, conclusions, or predictions from observed cases, including cases with incomplete information.
- How it is used
- Uses observed data to expand knowledge bases and support predictions or actions when cases are incomplete or previously unseen.
- How deep it goes
- Explains generalization, uncertainty, and a structured-data pipeline, but stops short of detailed learning algorithms or mathematical derivations.
- Examples
- Uses spam classification, coin and grandfather generalizations, and graph embeddings for social, biological, and medical networks.
- What you can do after
- Evaluate/understand how models generalize from observations, structure raw data, and make predictions despite incomplete information.
Read how Knowledge Graphs and LLMs in Action explains inductive inference →
Graph Neural Networks in Action
Inductive learning learns a generalizable function or representation from observed data for application to unseen data.
- How it is used
- Supports embedding and prediction for unseen nodes or graphs when graph data changes.
- How deep it goes
- Explains the learned feature-to-embedding mapping and contrasts it with fixed-graph embeddings, but does not give an inductive implementation.
- Examples
- Uses unseen nodes, new graphs, and evolving networks to contrast GNN embeddings with Node2Vec embeddings.
- What you can do after
- Evaluate/understand when to choose GNN embeddings over Node2Vec for changing graphs and feature-rich tasks.
Read how Graph Neural Networks in Action explains inductive inference →
Sutskever's List
Inductive inference is the process of inferring the process or hypothesis that could have generated observed data.
- How it is used
- Uses inductive inference to ground an idealized agent intended to move beyond systems built for specific tasks.
- How deep it goes
- Defines the problem and illustrates two philosophical constraints, then motivates weighted consideration of computable programs without deriving a formal method.
- Examples
- Uses missing cookies with competing explanations and the sequence 1, 3, 5, 7… to contrast compatible hypotheses and simplicity.
- What you can do after
- Understand why inference should retain data-compatible hypotheses while favoring simpler explanations.
Related concepts
- transductive learningcontrasts with · Transductive learning infers representations or predictions for specific entities and structures observed during training rather than learning rules that generalize to unseen entities.
- generalizationproduces · Generalization is the extension of a pattern, model, solution, or conclusion beyond the specific examples or cases from which it was derived.
- multi-step reasoningkind of · Multi-step reasoning derives a conclusion, solution, or decision through multiple connected intermediate inferential or problem-solving steps.
- deductive reasoningcontrasts with · Deductive reasoning is the process of deriving conclusions that necessarily follow from stated general premises, rules, or hypotheses.
- machine learninguses · Machine learning is a field of computational methods in which systems learn patterns or relationships from data to make predictions, decisions, or other outputs.
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.