Genetic algorithm for neural network architecture optimization

Idrissi, M. A. Janati; Ramchoun, Hassan; Ghanou, Youssef; Ettaouil, Mohamed

doi:10.1109/gol.2016.7731699

Cited by 46 publications

(23 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Figure 6 shows the parameter adjustment of the chromosome. The control genes are adjusted using a GA [20,21], whereas the parameter genes are tuned using PSO. The GA is a stochastic search procedure based on natural selection with the use of reproduce, crossover, and mutation operations.…”

Section: Local Search Using Chaotic Methodsmentioning

confidence: 99%

Unstable System Control Using an Improved Particle Swarm Optimization-Based Neural Network Controller

Lin

Jhang

2019

Electronics

View full text Add to dashboard Cite

In this study, an improved particle swarm optimization (IPSO)-based neural network controller (NNC) is proposed for solving a real unstable control problem. The proposed IPSO automatically determines an NNC structure by a hierarchical approach and optimizes the parameters of the NNC by chaos particle swarm optimization. The proposed NNC based on an IPSO learning algorithm is used for controlling a practical planetary train-type inverted pendulum system. Experimental results show that the robustness and effectiveness of the proposed NNC based on IPSO are superior to those of other methods.

show abstract

Section: Local Search Using Chaotic Methodsmentioning

confidence: 99%

Unstable System Control Using an Improved Particle Swarm Optimization-Based Neural Network Controller

Lin

Jhang

2019

Electronics

View full text Add to dashboard Cite

show abstract

“…El artículo consultado (Idrissi, Ramchoun, Ghanou y Ettaouil, 2016) propone una formulación matemática multiobjetivo, a fin de determinar el número óptimo de capas ocultas, el número de neuronas en cada capa y el mejor valor de los pesos de sus respectivas salidas. Por lo general, la elección de la arquitectura de la RNA es hecha de forma empírica.…”

Section: Genetic Algorithm For Neural Network Architecture Optimizationunclassified

Prediccion de desercion laboral utilizando un algoritmo genetico y redes neuronales artificiales

Reyes-Huertas

2019

Interfases

View full text Add to dashboard Cite

Predicting job abandonment through genetic algorithms and artificial neural networks This research work aims to develop a tool to identify employees who might abandon their position, because job abandonment is considered an international problem. The proposed method consists of a genetic algorithm that allows identifying the significant variables and improving the architecture of an artificial neural network as a solution. The variables selected by the tool were similar to those collected from different studies but not all of them were considered in such studies (e.g., distance between home and workplace, and years of employment). Likewise, the variables and architecture selected by the tool allowed to predict job abandonment up to 88.92 % accuracy rate.

show abstract

“…7 while generation ≤ generation max do 8 if generation = 0 then 9 epochs ← (epochs +1). 10 population size ← (population size −10). 11 Select the parents.…”

Section: Genetic Algorithmmentioning

confidence: 99%

“…• number of filters in 1D and 2D convolution: [10,100] • filter size for 1D and 2D convolution: [1,6] • kernel size for 1D and 2D max pooling: [1,6] • number of units in fully connected layers: [ new layer ← CreateLayer(dropout allowed, layer type) 10 Append newlayer to chromosome 11 if newlayer is fully connected then 12 layer type ← 'non-cnn' 13 Randomly create fully connected layer and append to chromosome 14 return chromosome. 15 Function CreateLayer (dropout, type) 16 if type = 'cnn then 17 if dropout = true then 18 Randomly create convolution, fully connected or dropout operation 19 else 20 Randomly create convolution layer 21 else 22 if dropout = true then 23 Randomly create fully connected layer or dropout operation 24 else 25 Randomly create fully connected layer…”

Section: A Initial Population Generationmentioning

confidence: 99%

EDEN: Evolutionary deep networks for efficient machine learning

Dufourq

Bassett

2017

2017 Pattern Recognition Association of South Africa and Robotics and Mechatronics (PRASA-RobMech)

View full text Add to dashboard Cite

Deep neural networks continue to show improved performance with increasing depth, an encouraging trend that implies an explosion in the possible permutations of network architectures and hyperparameters for which there is little intuitive guidance. To address this increasing complexity, we propose Evolutionary DEep Networks (EDEN), a computationally efficient neuro-evolutionary algorithm which interfaces to any deep neural network platform, such as TensorFlow. We show that EDEN evolves simple yet successful architectures built from embedding, 1D and 2D convolutional, max pooling and fully connected layers along with their hyperparameters. Evaluation of EDEN across seven image and sentiment classification datasets shows that it reliably finds good networks -and in three cases achieves state-of-the-art results -even on a single GPU, in just 6-24 hours. Our study provides a first attempt at applying neuro-evolution to the creation of 1D convolutional networks for sentiment analysis including the optimisation of the embedding layer.

show abstract

Genetic algorithm for neural network architecture optimization

Cited by 46 publications

References 6 publications

Unstable System Control Using an Improved Particle Swarm Optimization-Based Neural Network Controller

Unstable System Control Using an Improved Particle Swarm Optimization-Based Neural Network Controller

Prediccion de desercion laboral utilizando un algoritmo genetico y redes neuronales artificiales

EDEN: Evolutionary deep networks for efficient machine learning

Contact Info

Product

Resources

About