TRAINING NEURAL NETWORKS BY MEANS OF GENETIC ALGORITHMS WORKING ON VERY LONG CHROMOSOMES

1995 ◽  
Vol 06 (03) ◽  
pp. 299-316 ◽  
Author(s):  
PETER G. KORNING
Keyword(s):  
Genetic Algorithm ◽  
Neural Networks ◽  
Genetic Algorithms ◽  
Fitness Function ◽  
Multiple Representations ◽  
Real Life ◽  
Multilayer Perceptrons ◽  
Training Methods ◽  
Neural Net ◽  
The Neural Network

In the neural network/genetic algorithm community, rather limited success in the training of neural networks by genetic algorithms has been reported. In a paper by Whitley et al. (1991), he claims that, due to “the multiple representations problem”, genetic algorithms will not effectively be able to train multilayer perceptrons, whose chromosomal representation of its weights exceeds 300 bits. In the following paper, by use of a “real-life problem”, known to be non-trivial, and by a comparison with “classic” neural net training methods, I will try to show, that the modest success of applying genetic algorithms to the training of perceptrons, is caused not so much by the “multiple representations problem” as by the fact that problem-specific knowledge available is often ignored, thus making the problem unnecessarily tough for the genetic algorithm to solve. Special success is obtained by the use of a new fitness function, which takes into account the fact that the search performed by a genetic algorithm is holistic, and not local as is usually the case when perceptrons are trained by traditional methods.

scholarly journals Training of Neural Networks by means of Genetic Algorithms Working on very long Chromosomes

1994 ◽  
Vol 23 (486) ◽  
Author(s):  
Peter Korning
Keyword(s):  
Genetic Algorithm ◽  
Neural Networks ◽  
Genetic Algorithms ◽  
Multiple Representations ◽  
Real Life ◽  
Multilayer Perceptrons ◽  
Training Methods ◽  
Neural Net ◽  
Specific Knowledge ◽  
The Neural Network

In the neural network / genetic algorithm community, rather limited success in the training of neural networks by genetic algorithms has been reported. In a paper by Whitley (1991), he claims that, due to ``the multiple representations problem´´, genetic algorithms will not effectively be able to train multilayer perceptrons, whose chromosomal representation of its weights exceeds 300 bit's. In the following paper, by use of a ``real-life problem´´, known to be non-trivial, and by a comparison with ``classic´´ neural net training methods, I will try to show that the modest success of applying genetic algorithms to the training of perceptrons, is caused not so much by the ``multiple representations problem´´ as by the fact that problem-specific knowledge available is often ignored, thus making the problem unnecessarily tough for the genetic algorithm to solve.

scholarly journals NEAT Snake: a both evolutionary and neural network adaptation approach

2020 ◽  
Author(s):  
Alisson Steffens Henrique ◽  
Vinicius Almeida dos Santos ◽  
Rodrigo Lyra
Keyword(s):  
Neural Network ◽  
Genetic Algorithm ◽  
Neural Networks ◽  
Decision Making ◽  
Genetic Algorithms ◽  
Real Time ◽  
Relative Position ◽  
Network Adaptation ◽  
The Neural Network

There are several challenges when modeling artificial intelligencemethods for autonomous players on games (bots). NEAT is one ofthe models that, combining genetic algorithms and neural networks,seek to describe a bot behavior more intelligently. In NEAT, a neuralnetwork is used for decision making, taking relevant inputs fromthe environment and giving real-time decisions. In a more abstractway, a genetic algorithm is applied for the learning step of the neuralnetworks’ weights, layers, and parameters. This paper proposes theuse of relative position as the input of the neural network, basedon the hypothesis that the bot profit will be improved.

scholarly journals The Use of Neural Networks and Genetic Algorithms to Control Low Rigidity Shafts Machining

Sensors ◽  
2020 ◽  
Vol 20 (17) ◽  
pp. 4683
Author(s):  
Antoni Świć ◽  
Dariusz Wołos ◽  
Arkadiusz Gola ◽  
Grzegorz Kłosowski
Keyword(s):  
Neural Network ◽  
Machine Learning ◽  
Genetic Algorithm ◽  
Neural Networks ◽  
Genetic Algorithms ◽  
Short Term Memory ◽  
Real Life ◽  
Control Process ◽  
Lstm Network ◽  
High Level

The article presents an original machine-learning-based automated approach for controlling the process of machining of low-rigidity shafts using artificial intelligence methods. Three models of hybrid controllers based on different types of neural networks and genetic algorithms were developed. In this study, an objective function optimized by a genetic algorithm was replaced with a neural network trained on real-life data. The task of the genetic algorithm is to select the optimal values of the input parameters of a neural network to ensure minimum deviation. Both input vector values and the neural network’s output values are real numbers, which means the problem under consideration is regressive. The performance of three types of neural networks was analyzed: a classic multilayer perceptron network, a nonlinear autoregressive network with exogenous input (NARX) prediction network, and a deep recurrent long short-term memory (LSTM) network. Algorithmic machine learning methods were used to achieve a high level of automation of the control process. By training the network on data from real measurements, we were able to control the reliability of the turning process, taking into account many factors that are usually overlooked during mathematical modelling. Positive results of the experiments confirm the effectiveness of the proposed method for controlling low-rigidity shaft turning.

scholarly journals Application of Genetic Algorithm Elements to Modelling of Rotation Processes in Motion Transmission Including a Long Shaft

Energies ◽  
2020 ◽  
Vol 14 (1) ◽  
pp. 115
Author(s):  
Andriy Chaban ◽  
Marek Lis ◽  
Andrzej Szafraniec ◽  
Radoslaw Jedynak
Keyword(s):  
Genetic Algorithm ◽  
Genetic Algorithms ◽  
Fitness Function ◽  
Least Square ◽  
Distributed Parameter ◽  
Parameter Method ◽  
Analytical Mechanics ◽  
Parameter System ◽  
Rotational Systems ◽  
Elastic Elements

Genetic algorithms are used to parameter identification of the model of oscillatory processes in complicated motion transmission of electric drives containing long elastic shafts as systems of distributed mechanical parameters. Shaft equations are generated on the basis of a modified Hamilton–Ostrogradski principle, which serves as the foundation to analyse the lumped parameter system and distributed parameter system. They serve to compute basic functions of analytical mechanics of velocity continuum and rotational angles of shaft elements. It is demonstrated that the application of the distributed parameter method to multi-mass rotational systems, that contain long elastic elements and complicated control systems, is not always possible. The genetic algorithm is applied to determine the coefficients of approximation the system of Rotational Transmission with Elastic Shaft by equivalent differential equations. The fitness function is determined as least-square error. The obtained results confirm that application of the genetic algorithms allow one to replace the use of a complicated distributed parameter model of mechanical system by a considerably simpler model, and to eliminate sophisticated calculation procedures and identification of boundary conditions for wave motion equations of long elastic elements.

scholarly journals Strategic Team AI Path Plans: Probabilistic Pathfinding

2008 ◽  
Vol 2008 ◽  
pp. 1-6 ◽  
Author(s):  
Tng C. H. John ◽  
Edmond C. Prakash ◽  
Narendra S. Chaudhari
Keyword(s):  
Genetic Algorithm ◽  
Genetic Algorithms ◽  
Computer Games ◽  
Fitness Function ◽  
New Method ◽  
High Quality ◽  
Plan Generation ◽  
Novel Method ◽  
Games And Simulations

This paper proposes a novel method to generate strategic team AI pathfinding plans for computer games and simulations using probabilistic pathfinding. This method is inspired by genetic algorithms (Russell and Norvig, 2002), in that, a fitness function is used to test the quality of the path plans. The method generates high-quality path plans by eliminating the low-quality ones. The path plans are generated by probabilistic pathfinding, and the elimination is done by a fitness test of the path plans. This path plan generation method has the ability to generate variation or different high-quality paths, which is desired for games to increase replay values. This work is an extension of our earlier work on team AI: probabilistic pathfinding (John et al., 2006). We explore ways to combine probabilistic pathfinding and genetic algorithm to create a new method to generate strategic team AI pathfinding plans.

scholarly journals Application of the genetic algorithm to the task of compiling the curriculum

2018 ◽  
Author(s):  
V. A. Turchina ◽  
D. O. Tanasienko
Keyword(s):  
Higher Education ◽  
Genetic Algorithm ◽  
Genetic Algorithms ◽  
Necessary Conditions ◽  
Fitness Function ◽  
Educational Process ◽  
Heuristic Methods ◽  
Educational Groups ◽  
The Individual ◽  
Negative Property

One of the main tasks in organizing the educational process in higher education is the drawing up of a schedule of classes. It reflects the weekly student and faculty load. At the same time, when compiling, there are a number of necessary conditions and a number of desirable. The paper considers seven required and four desirable conditions. In this paper, one of the well-known approaches that can be used in drawing up a curriculum is consid-ered. The proposed scheme of the genetic algorithm, the result of which is to obtain an approximate solution to the problem of scheduling with the need to further improve it by other heuristic methods. To solve the problem, an island model of the genetic algorithm was selected and its advantages were considered. In the paper, the author's own structure of the individual, which includes chromosomes in the form of educational groups and genes as a lesson at a certain time, is presented and justified. The author presents his own implementations of the genetic algorithms. During the work, many variants of operators were tested, but they were rejected due to their inefficiency. The biggest problem was to maintain the consistency of information encoded in chromosomes. Also, two post-steps were added: to try to reduce the number of teacher conflict conflicts and to normalize the schedule - to remove windows from the schedule. The fitness function is calculated according to the following principles: if some desired or desired property is present in the individual, then a certain number is deducted from the individual's assessment, if there is a negative property, then a certain number is added to the assessment. Each criterion has its weight, so the size of the fine or rewards may be different. In this work, fines were charged for non-fulfillment of mandatory conditions, and rewards for fulfilling the desired

scholarly journals PENENTUAN PARAMETER LEARNING RATE SELAMA PEMBELAJARAN JARINGAN SYARAF TIRUAN BACKPROPAGATION MENGGUNAKAN ALGORITMA GENETIKA

2020 ◽  
Vol 14 (2) ◽  
pp. 202-212
Author(s):  
Ade chandra Saputra
Keyword(s):  
Neural Network ◽  
Genetic Algorithm ◽  
Genetic Algorithms ◽  
Fitness Function ◽  
Learning Rate ◽  
Local Minima ◽  
Rate Parameter ◽  
Neural Network Learning ◽  
Value Of Learning ◽  
Backpropagation Learning

One of the weakness in backpropagation Artificial neural network(ANN) is being stuck in local minima. Learning rate parameter is an important parameter in order to determine how fast the ANN Learning. This research is conducted to determine a method of finding the value of learning rate parameter using a genetic algorithm when neural network learning stops and the error value is not reached the stopping criteria or has not reached the convergence. Genetic algorithm is used to determine the value of learning rate used is based on the calculation of the fitness function with the input of the ANN weights, gradient error, and bias. The calculation of the fitness function will produce an error value of each learning rate which represents each candidate solutions or individual genetic algorithms. Each individual is determined by sum of squared error value. One with the smallest SSE is the best individual. The value of learning rate has chosen will be used to continue learning so that it can lower the value of the error or speed up the learning towards convergence. The final result of this study is to provide a new solution to resolve the problem in the backpropagation learning that often have problems in determining the learning parameters. These results indicate that the method of genetic algorithms can provide a solution for backpropagation learning in order to decrease the value of SSE when learning of ANN has been static in large error conditions, or stuck in local minima

A Genetic Algorithm to Improve a Neural Network to Predict a Patient’s Response to Warfarin

1993 ◽  
Vol 32 (01) ◽  
pp. 55-58 ◽  
Author(s):  
M. N. Narayanan ◽  
S. B. Lucas
Keyword(s):  
Neural Network ◽  
Genetic Algorithm ◽  
Neural Networks ◽  
Predictive Ability ◽  
Predictor Variables ◽  
Optimal Subset ◽  
The Neural Network ◽  
International Normalised Ratio ◽  
The Mean ◽  
Paired T Test

Abstract:The ability of neural networks to predict the international normalised ratio (INR) for patients treated with Warfarin was investigated. Neural networks were obtained by using all the predictor variables in the neural network, or by using a genetic algorithm to select an optimal subset of predictor variables in a neural network. The use of a genetic algorithm gave a marked and significant improvement in the prediction of the INR in two of the three cases investigated. The mean error in these cases, typically, reduced from 1.02 ± 0.29 to 0.28 ± 0.25 (paired t-test, t = −4.71, p <0.001, n = 30). The use of a genetic algorithm with Warfarin data offers a significant enhancement of the predictive ability of a neural network with Warfarin data, identifies significant predictor variables, reduces the size of the neural network and thus the speed at which the reduced network can be trained, and reduces the sensitivity of a network to over-training.

scholarly journals The Use of Neural Networks and Genetic Algorithms for Design of Groundwater Remediation Schemes

1997 ◽  
Vol 1 (2) ◽  
pp. 345-356 ◽  
Author(s):  
Z. Rao ◽  
D. G. Jamieson
Keyword(s):  
Neural Network ◽  
Genetic Algorithm ◽  
Neural Networks ◽  
Artificial Neural Network ◽  
Genetic Algorithms ◽  
Hybrid Algorithm ◽  
Large Scale ◽  
Groundwater Remediation ◽  
Injection Wells ◽  
Representative Problem

Abstract. The increasing incidence of groundwater pollution has led to recognition of a need to develop objective techniques for designing reniediation schemes. This paper outlines one such possibility for determining how many abstraction/injection wells are required, where they should be located etc., having regard to minimising the overall cost. To that end, an artificial neural network is used in association with a 2-D or 3-D groundwater simulation model to determine the performance of different combinations of abstraction/injection wells. Thereafter, a genetic algorithm is used to identify which of these combinations offers the least-cost solution to achieve the prescribed residual levels of pollutant within whatever timescale is specified. The resultant hybrid algorithm has been shown to be effective for a simplified but nevertheless representative problem; based on the results presented, it is expected the methodology developed will be equally applicable to large-scale, real-world situations.

scholarly journals Maximization of Eigenfrequency Gaps in a Composite Cylindrical Shell Using Genetic Algorithms and Neural Networks

2019 ◽  
Vol 9 (13) ◽  
pp. 2754 ◽  
Author(s):  
Bartosz Miller ◽  
Leonard Ziemiański
Keyword(s):  
Genetic Algorithm ◽  
Neural Networks ◽  
Genetic Algorithms ◽  
Excitation Frequency ◽  
Algorithm Optimization ◽  
External Excitation ◽  
Excitation Frequencies ◽  
Objective Function Values ◽  
Using Data ◽  
Laminated Structures

This paper presents a novel method for the maximization of eigenfrequency gaps around external excitation frequencies by stacking sequence optimization in laminated structures. The proposed procedure enables the creation of an array of suggested lamination angles to avoid resonance for each excitation frequency within the considered range. The proposed optimization algorithm, which involves genetic algorithms, artificial neural networks, and iterative retraining of the networks using data obtained from tentative optimization loops, is accurate, robust, and significantly faster than typical genetic algorithm optimization in which the objective function values are calculated using the finite element method. The combined genetic algorithm–neural network procedure was successfully applied to problems related to the avoidance of vibration resonance, which is a major concern for every structure subjected to periodic external excitations. The presented examples illustrate a combined approach to avoiding resonance through the maximization of a frequency gap around external excitation frequencies complemented by the maximization of the fundamental natural frequency. The necessary changes in natural frequencies are caused only by appropriate changes in the lamination angles. The investigated structures are thin-walled, laminated one- or three-segment shells with different boundary conditions.

Sign in / Sign up

Export Citation Format

Share Document