Predicting Aflatoxin Contamination in Peanuts: A Genetic Algorithm/Neural Network Approach

Authors:
C. E. Henderson;W. D. Potter;R. W. McClendon;G. Hoogenboom
Affiliations:
Artificial Intelligence Center, University of Georgia, USA;Artificial Intelligence Center, University of Georgia, USA;Biological and Agricultural Engineering, University of Georgia, USA;Biological and Agricultural Engineering, University of Georgia, USA
Venue:
Applied Intelligence
Year:
2000

Citing 7
Cited 2

Towards the genetic synthesis of neural network

Proceedings of the third international conference on Genetic algorithms
Designing neural networks using genetic algorithms

Proceedings of the third international conference on Genetic algorithms
A general framework for parallel distributed processing

Parallel distributed processing: explorations in the microstructure of cognition, vol. 1
Using genetic search to exploit the emergent behavior of neural networks

Emergent computation
An introduction to genetic algorithms

An introduction to genetic algorithms
Genetic Algorithms in Search, Optimization and Machine Learning

Genetic Algorithms in Search, Optimization and Machine Learning
A comparison between cellular encoding and direct encoding for genetic neural networks

GECCO '96 Proceedings of the 1st annual conference on Genetic and evolutionary computation

An optimization methodology for machine learning strategies and regression problems in ballistic impact scenarios

Applied Intelligence
Simultaneous optimization of artificial neural networks for financial forecasting

Applied Intelligence

Quantified Score

Hi-index	0.00

Visualization

Abstract

Aflatoxin contamination in peanut crops is a problemof significant health and financial importance. Predicting aflatoxinlevels prior to crop harvest is useful for minimizing the impact of acontaminated crop and is the goal of our research. Backpropagationneural networks have been used to model problems of this type,however development of networks poses the complex problem of settingvalues for architectural features and backpropagationparameters. Genetic algorithms have been used in other studies todetermine parameters for backpropagation neural networks. This paperdescribes the development of a genetic algorithm/backpropagationneural network hybrid (GA/BPN) in which a genetic algorithm is usedto find architectures and backpropagation parameter valuessimultaneously for a backpropagation neural network that predictsaflatoxin contamination levels in peanuts based on environmentaldata. Learning rate, momentum, and number of hidden nodes are theparameters that are set by the genetic algorithm. A three-layerfeed-forward network with logistic activation functions is used.Inputs to the network are soil temperature, drought duration, cropage, and accumulated heat units. The project showed that the GA/BPNapproach automatically finds highly fit parameter sets forbackpropagation neural networks for the aflatoxin problem.