dorsal/arxiv
View SchemaNEVO-GSPT: Population-Based Neural Network Evolution Using Inflate and Deflate Operators
| Authors | Davide Farinati, Frederico J. J. B. Santos, Leonardo Vanneschi, Mauro Castelli |
|---|---|
| Categories | |
| ArXiv ID | 2601.08657vv1 |
| URL | https://arxiv.org/abs/2601.08657 |
| License | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ |
Abstract
Evolving neural network architectures is a computationally demanding process. Traditional methods often require an extensive search through large architectural spaces and offer limited understanding of how structural modifications influence model behavior. This paper introduces \gls{ngspt}, a novel Neuroevolution algorithm based on two key innovations. First, we adapt geometric semantic operators~(GSOs) from genetic programming to neural network evolution, ensuring that architectural changes produce predictable effects on network semantics within a unimodal error surface. Second, we introduce a novel operator (DGSM) that enables controlled reduction of network size, while maintaining the semantic properties of~GSOs. Unlike traditional approaches, \gls{ngspt}'s efficient evaluation mechanism, which only requires computing the semantics of newly added components, allows for efficient population-based training, resulting in a comprehensive exploration of the search space at a fraction of the computational cost. Experimental results on four regression benchmarks show that \gls{ngspt} consistently evolves compact neural networks that achieve performance comparable to or better than established methods in the literature, such as standard neural networks, SLIM-GSGP, TensorNEAT, and SLM.
{
"annotation_id": "955929be-dfe8-44c1-8e16-9b78283ac2b3",
"date_created": "2026-02-17T05:53:15.696000Z",
"date_modified": "2026-02-17T05:53:15.696000Z",
"file_hash": "01a9daeb95e43552cbb5df9d12d0271b77c6e765185932758b111d590f08b8c9",
"private": false,
"record": {
"abstract": "Evolving neural network architectures is a computationally demanding process. Traditional methods often require an extensive search through large architectural spaces and offer limited understanding of how structural modifications influence model behavior. This paper introduces \\gls{ngspt}, a novel Neuroevolution algorithm based on two key innovations. First, we adapt geometric semantic operators~(GSOs) from genetic programming to neural network evolution, ensuring that architectural changes produce predictable effects on network semantics within a unimodal error surface. Second, we introduce a novel operator (DGSM) that enables controlled reduction of network size, while maintaining the semantic properties of~GSOs. Unlike traditional approaches, \\gls{ngspt}\u0027s efficient evaluation mechanism, which only requires computing the semantics of newly added components, allows for efficient population-based training, resulting in a comprehensive exploration of the search space at a fraction of the computational cost. Experimental results on four regression benchmarks show that \\gls{ngspt} consistently evolves compact neural networks that achieve performance comparable to or better than established methods in the literature, such as standard neural networks, SLIM-GSGP, TensorNEAT, and SLM.",
"arxiv_id": "2601.08657",
"authors": [
"Davide Farinati",
"Frederico J. J. B. Santos",
"Leonardo Vanneschi",
"Mauro Castelli"
],
"categories": [
"cs.NE"
],
"license": "http://arxiv.org/licenses/nonexclusive-distrib/1.0/",
"title": "NEVO-GSPT: Population-Based Neural Network Evolution Using Inflate and Deflate Operators",
"url": "https://arxiv.org/abs/2601.08657",
"version": "v1"
},
"schema_id": "dorsal/arxiv",
"source": {
"execution_id": "ba60765a-6cf6-4043-8c7a-eea1b5bafa62",
"id": "arXiv Dataset",
"type": "Model",
"variant": "snapshot-2026-01-17",
"version": "0.1.0"
},
"user_id": 1000002
}