Cascade-Correlation learning architecture
E474908
Cascade-Correlation learning architecture is a neural network training method that incrementally builds its own topology by adding new hidden units during learning to improve performance.
All labels observed (5)
How this entity was disambiguated
This entity first appeared as the object of triple T4850106 — resolving that mention is where its identity was fixed. The disambiguator weighed these candidate entities and picked the highlighted one (or “None”, minting a new entity). This is how homonymy is resolved: the same surface form can point to different entities.
NED1
Entity disambiguation (via context triple)
gpt-5-mini-2025-08-07
Target entity: Cascade-Correlation learning architecture Context triple: [Scott Fahlman, knownFor, Cascade-Correlation learning architecture]
-
A.
Hopfield networks
Hopfield networks are recurrent artificial neural networks that serve as content-addressable memory systems, storing patterns as stable states and retrieving them through dynamics that minimize an energy function.
-
B.
Gradient-based learning applied to document recognition
"Gradient-based learning applied to document recognition" is a seminal 1998 paper by Yann LeCun and colleagues that introduced and demonstrated the effectiveness of convolutional neural networks for tasks like handwritten digit recognition, helping to lay the foundations of modern deep learning.
-
C.
“Learning representations by back-propagating errors”
“Learning representations by back-propagating errors” is a landmark 1986 research paper that popularized the backpropagation algorithm for training multi-layer neural networks, helping to launch the modern field of deep learning.
-
D.
Helmholtz machine
The Helmholtz machine is a pioneering generative neural network model that learns internal representations by using separate recognition and generative pathways to perform unsupervised learning.
-
E.
“A fast learning algorithm for deep belief nets”
“A fast learning algorithm for deep belief nets” is a seminal 2006 paper by Geoffrey Hinton that introduced an efficient unsupervised pretraining method for deep neural networks using stacked restricted Boltzmann machines.
- F. None of above. chosen
- G. Unsure - the case is ambiguous/there is not enough information to decide.
NED2
Entity disambiguation (via description)
gpt-5-mini-2025-08-07
Target entity: Cascade-Correlation learning architecture Target entity description: Cascade-Correlation learning architecture is a neural network training method that incrementally builds its own topology by adding new hidden units during learning to improve performance.
-
A.
Hopfield networks
Hopfield networks are recurrent artificial neural networks that serve as content-addressable memory systems, storing patterns as stable states and retrieving them through dynamics that minimize an energy function.
-
B.
Gradient-based learning applied to document recognition
"Gradient-based learning applied to document recognition" is a seminal 1998 paper by Yann LeCun and colleagues that introduced and demonstrated the effectiveness of convolutional neural networks for tasks like handwritten digit recognition, helping to lay the foundations of modern deep learning.
-
C.
“Learning representations by back-propagating errors”
“Learning representations by back-propagating errors” is a landmark 1986 research paper that popularized the backpropagation algorithm for training multi-layer neural networks, helping to launch the modern field of deep learning.
-
D.
Helmholtz machine
The Helmholtz machine is a pioneering generative neural network model that learns internal representations by using separate recognition and generative pathways to perform unsupervised learning.
-
E.
“A fast learning algorithm for deep belief nets”
“A fast learning algorithm for deep belief nets” is a seminal 2006 paper by Geoffrey Hinton that introduced an efficient unsupervised pretraining method for deep neural networks using stacked restricted Boltzmann machines.
- F. None of above. chosen
Statements (46)
| Predicate | Object |
|---|---|
| instanceOf |
constructive neural network algorithm
ⓘ
neural network training method ⓘ supervised learning algorithm ⓘ |
| abbreviation |
CasCor
ⓘ
Cascade-Correlation ⓘ
linked to:
Cascade-Correlation learning architecture
|
| addsHiddenUnits | one at a time ⓘ |
| architectureProperty | topology determined during training rather than fixed a priori ⓘ |
| category |
adaptive network architecture method
ⓘ
constructive learning algorithm ⓘ |
| comparedWith | backpropagation ⓘ |
| describedIn |
The Cascade-Correlation Learning Architecture
ⓘ
linked to:
Cascade-Correlation learning architecture
|
| designedToImprove |
generalization performance
ⓘ
training speed compared to standard backpropagation ⓘ |
| developedBy |
Christian Lebiere
ⓘ
Scott E. Fahlman ⓘ |
| field |
machine learning
ⓘ
neural networks ⓘ |
| hasAuthor |
Christian Lebiere
ⓘ
Scott E. Fahlman ⓘ |
| hasKeyIdea |
adds new hidden units during training
ⓘ
constructive growth of network architecture ⓘ freezes weights of previously learned units ⓘ incrementally builds its own network topology ⓘ new hidden units are trained to maximize correlation with residual error ⓘ |
| hasLearningParadigm | supervised learning ⓘ |
| hasTrainingPhase |
candidate unit training phase
ⓘ
output weight training phase ⓘ |
| hiddenUnitConnectionPattern | new hidden units connect to all existing network units ⓘ |
| influenced |
constructive neural network methods
ⓘ
growing neural network architectures ⓘ |
| initialTopology | network starts with no hidden units ⓘ |
| introducedIn | 1990 ⓘ |
| languageOfOriginalPublication | English ⓘ |
| networkType | feedforward neural network ⓘ |
| optimizationTarget | correlation between candidate unit output and network residual error ⓘ |
| outputLayerTraining | output weights are retrained after adding each new hidden unit ⓘ |
| publicationVenue |
Advances in Neural Information Processing Systems
ⓘ
linked to:
NeurIPS
|
| stoppingCriterion | growth stops when performance no longer improves ⓘ |
| supports | incremental learning of network structure ⓘ |
| trainingObjective | reduce network error by adding hidden units ⓘ |
| usedFor |
classification
ⓘ
function approximation ⓘ regression ⓘ |
| usesOptimizationCriterion | maximization of correlation between unit output and network residual error ⓘ |
| usesWeightFreezing | true ⓘ |
| yearOfFirstPublication | 1990 ⓘ |
How these facts were elicited
The pipeline generated the facts above by prompting gpt-5.1 with this entity's name + description and the instruction below.
Instruction
You are a knowledge base construction expert. Given a subject entity and a description of it, return factual statements that you know for the subject as a JSON list of dictionaries(triples), where keys must be "subject", "predicate" and "object". The number of facts may be very high, between 25 to 50 or more, for very popular subjects. For less popular subjects, the number of facts can be very low, like 5 or 10. # Requirements - If you don't know the subject at all, return an empty list. - If the subject is not a named entity, return an empty list. - Include at least one triple where predicate is "instanceOf". - Do not get too wordy. - Separate several objects into multiple triples with one object.
Input
Subject: Cascade-Correlation learning architecture Description of subject: Cascade-Correlation learning architecture is a neural network training method that incrementally builds its own topology by adding new hidden units during learning to improve performance.
Referenced by (5)
Full triples — surface form annotated when it differs from this entity's canonical label.
linked to: Cascade-Correlation learning architecture
linked to: Cascade-Correlation learning architecture
Cascade-Correlation learning architecture
→
describedIn
→
The Cascade-Correlation Learning Architecture
ⓘ
linked to: Cascade-Correlation learning architecture
linked to: Cascade-Correlation learning architecture