Enum: NeuralNetworkPhenomenon
Training-time phenomena that affect neural-network learning, drawn from Clause 3.4.
URI: iso22989:NeuralNetworkPhenomenon
Permissible Values
| Value | Meaning | Description |
|---|---|---|
| vanishing_gradient | None | Gradient signal shrinks across layers during back-propagation, slowing learni... |
| exploding_gradient | None | Gradient signal grows without bound across layers during back-propagation (Cl... |
| catastrophic_forgetting | None | Previously learned knowledge is lost when the network is retrained on new dat... |
| overfitting | None | Model fits training data idiosyncrasies and fails to generalise |
| underfitting | None | Model lacks capacity or training to capture the underlying signal |
Slots
| Name | Description |
|---|---|
| training_phenomena | Training-time phenomena observed for the model (Clause 3 |
In Subsets
Identifier and Mapping Information
Schema Source
- from schema: https://w3id.org/lmodel/iso22989
LinkML Source
name: NeuralNetworkPhenomenon
description: Training-time phenomena that affect neural-network learning, drawn from
Clause 3.4.
in_subset:
- terminology
from_schema: https://w3id.org/lmodel/iso22989
rank: 1000
permissible_values:
vanishing_gradient:
text: vanishing_gradient
description: Gradient signal shrinks across layers during back-propagation, slowing
learning.
exploding_gradient:
text: exploding_gradient
description: Gradient signal grows without bound across layers during back-propagation
(Clause 3.4.5).
catastrophic_forgetting:
text: catastrophic_forgetting
description: Previously learned knowledge is lost when the network is retrained
on new data.
overfitting:
text: overfitting
description: Model fits training data idiosyncrasies and fails to generalise.
underfitting:
text: underfitting
description: Model lacks capacity or training to capture the underlying signal.