Evolving Genetic Programming Classifiers with Novelty Search
| hal.structure.identifier | Instituto Tecnológico de Tijuana = Tijuana Institute of Technology [Tijuana] | |
| dc.contributor.author | NAREDO, Enrique | |
| hal.structure.identifier | Instituto Tecnológico de Tijuana = Tijuana Institute of Technology [Tijuana] | |
| dc.contributor.author | LEONARDO, Trujillo | |
| hal.structure.identifier | Quality control and dynamic reliability [CQFD] | |
| hal.structure.identifier | Institut de Mathématiques de Bordeaux [IMB] | |
| dc.contributor.author | LEGRAND, Pierrick | |
| dc.date.accessioned | 2024-04-04T03:19:06Z | |
| dc.date.available | 2024-04-04T03:19:06Z | |
| dc.date.created | 2014-12-31 | |
| dc.identifier.uri | https://oskar-bordeaux.fr/handle/20.500.12278/194474 | |
| dc.description.abstractEn | Novelty Search (NS) is a unique approach towards search and optimization,where an explicit objective function is replaced by a measure of solution noveltyto provide the selective pressure in an artificial evolutionary system. However,NS has been mostly used in evolutionary robotics, while it’s applicability to classicmachine learning problems has been mostly unexplored. This work presents a NSbasedGenetic Programming (GP) algorithm for supervised classification, with thefollowing noteworthy contributions. It is shown that NS can solve real-world classificationtasks, validated over several commonly used benchmarks. These results aremade possible by using a domain-specific behavioral descriptor, closely related to theconcept of semantics in GP.Moreover, two new variants of the NS algorithm are proposed,Probabilistic NS (PNS) and a variant ofMinimum Criterion NS (MCNS). Theformer models the behavior of each solution as a random vector, eliminating all theNS parameters and reducing the computational overhead of the traditional NS algorithm;the latter uses a standard objective function to constrain the search and bias theprocess towards high performance solutions. The paper also discusses the effects ofNS on an important GP phenomenon, bloat. In particular, results indicate that somevariants of the NS approach can have a beneficial effect on the search process bycurtailing code growth. | |
| dc.language.iso | en | |
| dc.subject.en | Novelty Search | |
| dc.subject.en | Behavior-based Search | |
| dc.subject.en | Bloat | |
| dc.title.en | Evolving Genetic Programming Classifiers with Novelty Search | |
| dc.type | Document de travail - Pré-publication | |
| dc.subject.hal | Informatique [cs]/Intelligence artificielle [cs.AI] | |
| dc.subject.hal | Statistiques [stat]/Machine Learning [stat.ML] | |
| bordeaux.hal.laboratories | Institut de Mathématiques de Bordeaux (IMB) - UMR 5251 | * |
| bordeaux.institution | Université de Bordeaux | |
| bordeaux.institution | Bordeaux INP | |
| bordeaux.institution | CNRS | |
| hal.identifier | hal-01111234 | |
| hal.version | 1 | |
| hal.origin.link | https://hal.archives-ouvertes.fr//hal-01111234v1 | |
| bordeaux.COinS | ctx_ver=Z39.88-2004&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.au=NAREDO,%20Enrique&LEONARDO,%20Trujillo&LEGRAND,%20Pierrick&rft.genre=preprint |
Files in this item
| Files | Size | Format | View |
|---|---|---|---|
|
There are no files associated with this item. |
|||