Neuroevolution of Spiked Neural Networks With HyperNEAT
ISEF · 2024 Robotics and Intelligent Machines
Overview
Machine learning and neural networks have seen a rapid increase in popularity over the past few years. Spiked neuron networks (SNN) are a third generation neural network that offer many benefits compared to analog neural networks (ANN). While past work has used spikeprop or ANN conversion to train SNN, Daniel has researched an interesting alternative using evolutionary algorithms (EA) to train SNNs. This project uses Matlab to train SNNs with a specific EA called HyperNEAT in pursuit of creating a training method that can outperform more widely popular methods such as the ones listed above.
Competition history
- ISEF 2024
Resources
Related projects
ISEF · 2025
Evolving Neural Networks to Exploit Problem Processing Regularities
ISEF · 2022
Q-NEST: Quantum Neuroevolutionary Strategies for Parametrically Optimizing and Topologically Augmenting Artificial Neural Networks via Novel Mutation, Translocation, and Crossover Hybrid Quantum-Classical Genetic Operators
ISEF · 2025
Adapter-Enabled Multimodal Spiking Encoder/Decoder Neural Subnetwork Ensembles for Large-Scale Logical Neuromorphic Accelerators
ISEF · 2020
Implementing Supervised Deep Learning with Feedforward Neural Network Using Genetic Algorithms
ISEF · 2021
Neuroromorphic Computing: Simulating the Brain's Visual Cortex for a Faster, More Effecient Computer
ISEF · 2025
Analyzing the Optimization of Mechanical Neural Networks
ISEF · 2023
Improved Generalizability and Performance on Small Datasets Using Hyper-Connected Neural Architectures, C. elegans-Inspired Networks, and Evolutionary Modeling
ISEF · 2024
Autonomous Driving Agents Trained by an Algorithm Based on NeuroEvolution of Augmenting Topologies (Neat) Using a New Logarithmic Fixed-Point Number System Optimized for Microcontrollers
Closest projects by meaning, across every fair and year in the corpus.
Browse more like this
Source: Regeneron International Science and Engineering Fair