Genetic Algorithm Neural Network Model Research Articles

Classification for GM and Non-GM Maize Kernels Based on NIR Spectra and Deep Learning

The classification of GM and non-GM maize kernels is fundamental for further analysis of the gene action in maize. Therefore, a complete and novel detection scheme based on near-infrared spectra was designed to distinguish GM and non-GM maize kernels. Hyperspectral images (935–1720 nm) of 777 maize kernels from 3 kinds were captured, and the average spectra of the maize kernels were extracted for modeling analysis. The classical modeling methods based on feature engineering were first studied, and the backpropagation neural network–genetic algorithm model showed the best performance with a prediction accuracy of 0.861. Then, novel modeling methods based on deep learning were developed. To dig out the interactive information between different bands and match the application scenarios, the original spectra were transformed into two-dimensional matrices before establishing the deep learning models. A modified convolution neural network (i.e., VGG net) with dilated convolution was finally constructed to classify the maize kernels, and the prediction accuracy reached 0.961. This research provides a referential and novel way to detect GM maize kernels. Future research will improve the detection scheme for monitoring unauthorized GM organisms by introducing the visualization technology of deep learning.

Hybrid-biotaxonomy-like machine learning enables an anticipated surface plasmon resonance of Au/Ag nanoparticles assembled on ZnO nanorods

A genetic algorithm neural network model based on machine learning has been established for precise prediction of plasmonic resonance by considering the fabrication parameters of heterostructural noble metal nanoparticles.

Optimization of Enzymatic Hydrolysis of Soybean Dregs Based on Genetic Algorithm-Neural Network and Its Kinetics

Objective: The optimal process conditions for the solution of legumes by cellulase enzymes were obtained, and a model of the electrodynamics of bean slag enzymes was established. Methods: Based on the single-factor experiment, the effects of pH, cellulase addition, enzymatic temperature and enzyme solution time on reduced sugar production were studied by using the response surface method (RSM) and the genetic algorithm-neural network model (GA-ANN). Results: Genetic algorithm-neural network optimization of the bean slag enzyme solution process conditions were: pH5.2, cellulase addition 4.4%, enzymatic temperature 52 ℃, enzymatic time 3.2 h, under these conditions reduced sugar production was 2.65 g/kg, higher than the results of response surface optimization (reduced sugar production was 2.54 g/kg). The kinetics parameter k0 was 0.537 and the scropular dimensionality h was 0.153, according to the experimental data and the dynamic model, the fit was 0.9827, and the fitting effect was good. Conclusion: In this paper, the soybean slag enzyme solution process was optimized by using the response surface method and the genetic algorithm-neural network model, and the soybean slag enzyme solution dynamics model was established, which could provide theoretical reference for the soybean slag enzyme solution process.

Accurate prediction of the extrusion forming bonding reliability for heterogeneous welded sheets based on GA-BP neural network

Extrusion connection is a new method of forming and manufacturing heterogeneous welded sheets. The factors that affected the bonding quality are the forming temperature, the extrusion ratio, and the guiding angle of the die, which has brought trouble to the evaluation of bonding strength and quality. A method to establish a predicted model for the bonding strength of welded sheets by integrating finite element simulations, process experiments, and artificial neural networks was developed. Finite element simulations were used to verify the process experiments and provided training data sets for the artificial neural networks. The BP neural network was used to predict the bonding strength. Due to the randomness of the weight and threshold of the BP neural network, its predicted accuracy needs to be improved, in which genetic algorithms were used to optimize consequently. The results showed that the genetic algorithm neural network model had higher reliability, and the predicted accuracy was 99.5%. Compared with the traditional BP neural network, the predicted accuracy was improved by 5.78%, and the error was reduced to 0.5%. It has good generalization ability and provides a new way for intelligent reliability evaluation of high performance heterogeneous welded sheets via extrusion.

Study on the US Presidential Election Based on Neural Network

The world's most watched U.S. election is in a state of anxiety, with Republican candidate Donald John Trump and Democratic candidate Joe Biden will compete for the presidency this month. The two candidates will show different political positions and administrative options in different areas of development and domestic and foreign policy. The election of different candidates will have a great impact on the United States economy and the global economy, and will have a great impact on the trend of international relations. In response to question one, we need to analyze the possible impact of election of different candidates on the United States economy. We first quantitatively analyze the policy data of the two candidates obtained in the network in different fields. By merging the quantitative models of different fields, the AHP is used to calculate the weights of different fields, and the comparison between two candidates in each field and the impact on the American economy. According to the large data set collected, we do cluster analysis, and analyze and predict the impact of different fields by BP neural network model. For the second question, in order to analyze the impact of different candidate is elected to the Chinese economy, according to the problem a model for analysis, in order to improve the accuracy of the model, this part adds more considerations variables, at the same time, higher precision, based on genetic algorithm neural network model, after finishing the data concentration, 80% of the data used in the model, variables to determine index weight and threshold, the remaining 20% of the data by residual error calculation is used to model accuracy test. In view of question three, through the analysis of question one and question two and the actual situation, we can get the final winning candidate. And through the analysis of its economic and policy means to give China's economic countermeasures and policies in related fields. By referring to a large amount of data, this paper finally obtains the influence of different candidates on the united states, china and the global economy and politics through data visualization analysis of different candidates' policy claims in different fields. In order to further improve the model accuracy and computational efficiency, this paper also proposes to use Levenberg-Marquardt-based algorithm to solve the problem.

Multi-output Hybrid GA-NN with Adaptive Mechanism

This research presents a hybrid Genetic Algorithm Neural Network (GA-NN) model to simulate the physical tests procedures of Medium Density Fiberboard (MDF). Data included in the model are related to MDF properties and its fiber characteristics. Multilayer Perceptron NN is a reliable supervised machine learning model. The model learns from seven inputs fed to the network to produce prediction of three targets. In order to avoid result from local optimum scenario, GA optimizes synaptic weights of the network towards reducing prediction error. The research used a fixed probability rates for crossover and mutation for hybrid GA-NN model. GA-NN model is further improved using adaptive mechanism to help identify the best probability rates for crossover and mutation. Fitness value refers to Sum of Squared Error, which is the accumulation of network error in the Output Layer. The population fitness distribution will guide best rates for each epoch. Performance comparisons are among three models; namely NN with Back Propagation (BP), hybrid GA-NN and hybrid GA-NN with adaptive mechanism. Results show the hybrid GA-NN model perform much better than NN model with back propagation optimizer. Adaptive mechanism in GA helps increase capability to converge at zero sooner than the ordinary GA.

Soil salinity mapping by remote sensing south of Urmia Lake, Iran

Urmia Lake is a shallow terminal Lake located in northwest Iran and it is one of the largest permanent Lakes in the Middle East. In this study, the changes in soil salinity at Urmia Lake were investigated using satellite images and the oldest salinity map of the area over a period of 45 years from 1973 to 2018. The distribution of salinity in 2018 was estimated using the supervised classification by the nonlinear hybrid model of artificial multi-layered neural network-genetic algorithm model (ANN-GA) while the salinity map for the years of 1985, 1995, 2005 and 2015 was estimated by the unsupervised method. Further, the salinity data of surface soil in the region for the year 1973 was also digitized and utilized. For this purpose, 291 surface samples (258 samples for modeling and 33 samples for the re-evaluation of the model) of the studied region were collected and analyzed in 2018. The input neurons were selected by analyzing the satellite imagery bands, salinity indices, salinity ratio index and normalized difference vegetation index. The correlation coefficient and root-mean-square error of the training network model were equal to 0.94 and 0.04, respectively. The salinity map of the studied region was estimated using this model and classified into six classes (S0 to S5). The produced map of 2018 was used to re-evaluate the results. It showed that lower estimation accuracy was in classes S1 and S2. The obtained results in this study indicated that roughness, moisture, the density of halophyte plants and sodium slickspot were some of the sources for estimation of errors in lower salinity classes. The time-series changes in the salinity class of estimated maps showed that S3, S4 and S5 classes have expanded between 1973 and 2018. These are in agreement with the field observation and with the other scientific reports about the studied area.

High level extracellular production of recombinant human interferon alpha 2b in glycoengineered Pichia pastoris: culture medium optimization, high cell density cultivation and biological characterization.

The present study was aimed at design of experiments (DoE)- and artificial intelligence-based culture medium optimization for high level extracellular production of a novel recombinant human interferon alpha 2b (huIFNα2b) in glycoengineered Pichia pastoris and its characterization. The artificial neural network-genetic algorithm model exhibited improved huIFNα2b production and better predictability compared to response surface methodology. The optimized medium exhibited a fivefold increase in huIFNα2b titre compared to the complex medium. A maximum titre of huIFNα2b (436mgl-1 ) was achieved using the optimized medium in the bioreactor. Real-time capacitance data from dielectric spectroscopy were utilized to model the growth kinetics with unstructured models. Biological characterization by antiproliferative assay proved that the purified recombinant huIFNα2b was biologically active, exhibiting growth inhibition on breast cancer cell line. Culture medium optimization resulted in enhanced production of huIFNα2b in glycoengineered P. pastoris at both shake flask and bioreactor level. The purified huIFNα2b was found to be N-glycosylated and biologically active. DoE-based medium optimization strategy significantly improved huIFNα2b production. The antiproliferative activity of huIFNα2b substantiates its potential scope for application in cancer therapy.

Hybrid Design using Counter Propagation Neural Network-Genetic Algorithm Model for the Anomaly Detection in Online Transaction

In e-commerce, credit card fraud is an evolving challenge. The increase in the number of credit card transactions provides more opportunity for fraudsters to steal credit card numbers and execute fraud. Fraud detection is a continuously evolving discipline to tackle ever changing tactics to commit fraud. Existing fraud detection systems have not been so much efficient to reduce fraud transaction rate. Improvement in fraud detection practices has become essential to maintain existence of payment system. This research designed hybrid of Counter Propagation Neural Network and genetic algorithm (CPNN-GA) for the detection of anomaly in any online transactions.

Determining optimal combination of roller compacted concrete pavement mixture containing recycled asphalt pavement and crumb rubber using hybrid artificial neural network–genetic algorithm method considering energy absorbency approach

The present study investigates the effectiveness of evolutionary algorithms such as genetic algorithm (GA) evolved neural network in estimating roller compacted concrete pavement (RCCP) characteristics including flexural and compressive strength of RCC and also energy absorbency of mixes with different compositions. A real coded GA was implemented as training algorithm of feed forward neural network to simulate the models. The genetic operators were carefully selected to optimize the neural network, avoiding premature convergence and permutation problems. To evaluate the performance of the genetic algorithm neural network model, Nash-Sutcliffe efficiency criterion was employed and also utilized as fitness function for genetic algorithm which is a different approach for fitting in this area. The results showed that the GA-based neural network model gives a superior modeling. The well-trained neural network can be used as a useful tool for modeling RCC specifications.

Assessment of spatial distribution of soil heavy metals using ANN-GA, MSLR and satellite imagery.

This study aims to assess and compare heavy metal distribution models developed using stepwise multiple linear regression (MSLR) and neural network-genetic algorithm model (ANN-GA) based on satellite imagery. The source identification of heavy metals was also explored using local Moran index. Soil samples (n=300) were collected based on a grid and pH, organic matter, clay, iron oxide contents cadmium (Cd), lead (Pb) and zinc (Zn) concentrations were determined for each sample. Visible/near-infrared reflectance (VNIR) within the electromagnetic ranges of satellite imagery was applied to estimate heavy metal concentrations in the soil using MSLR and ANN-GA models. The models were evaluated and ANN-GA model demonstrated higher accuracy, and the autocorrelation results showed higher significant clusters of heavy metals around the industrial zone. The higher concentration of Cd, Pb and Zn was noted under industrial lands and irrigation farming in comparison to barren and dryland farming. Accumulation of industrial wastes in roads and streams was identified as main sources of pollution, and the concentration of soil heavy metals was reduced by increasing the distance from these sources. In comparison to MLSR, ANN-GA provided a more accurate indirect assessment of heavy metal concentrations in highly polluted soils. The clustering analysis provided reliable information about the spatial distribution of soil heavy metals and their sources.

Quantitative analysis of pit defects in an automobile engine cylinder cavity using the radial basis function neural network–genetic algorithm model

In the automotive remanufacturing movement, the inspection of the corrosion defects on the engine cylinder cavity is a key and difficult problem. In this article, based on the ultrasonic phased array technology and the radial basis function neural network–genetic algorithm model, a new quantitative analysis method is proposed to estimate the size of the pit defects on the automobile engine cylinder cavity. Echo signals from the small pit defects with different sizes are acquired by an ultrasonic phased array transducer. According to the ultrasonic signal characteristics, the feature vectors are extracted using wavelet packet, fractal technology, peak amplitude method, and some routine extract methods. The radial basis function neural network–genetic algorithm model is investigated for the quantitative analysis of the pit defects, which can obtain an optimal quantitative model. The results show that the proposed model is effective in the corrosion estimation work.

Appraisal of artificial neural network-genetic algorithm based model for prediction of the power provided by the agricultural tractors

The knowledge of the available power provided by the driving wheel of agricultural tractors is required to gain a correct insight into the energy management of agricultural tractors. The design of the tractors is pivotal on the maximization of the traction efficiency and simultaneous minimization of energy dissipation. This paper spearheads the synthesis of the power provided by the agricultural tractors as affected by wheel load, slip and speed by use of the potential of a soil bin facility and a single-wheel test rig. The hybridized artificial neural network-genetic algorithm method was adopted to model the provided power of the driving wheel under the effect of the aforementioned tire parameters. The common drawback of the back-propagation algorithm known as the low speed of convergence and the possibility of being trapped in a local minimum was solved by the use of genetic algorithm. The mean square error equal to 0.02242 was obtained as the most optimal artificial neural network-genetic algorithm configuration using Levenberg–Marquardt training algorithm. Therefore, a 3-9-1 feed-forward with back propagation learning algorithm was selected as the modeling structure. The computed coefficient of determination for the training and test phases of the best artificial neural network-genetic algorithm model was obtained at 0.9696 and 0.9672, respectively. The present study spearheads the required power estimation for the driving wheels of off-road vehicles while the experimental test conduction in a controlled soil bin facility using single-wheel tester and adoption of soft computing tools are of the highlights and added values of the paper.

Developing a hybrid artificial neural network-genetic algorithm model to predict resilient modulus of polypropylene/polyester fiber-reinforced asphalt concrete

Up to now various kinds of fibers are used to improve the hot mix asphalt (HMA) performance, but a few works have been undertaken on the hybrid fiber-reinforced HMA. Therefore, in this paper, the resilient modulus of the modified HMA samples using polypropylene and polyester fibers (hybrid and single modes) was evaluated and modeled by regression method and artificial neural network (ANN). As ANN includes different parameters such as the number of neurons in hidden layer influenced on the prediction accuracy, genetic algorithm (GA) was used to optimize the ANN parameters. Also, GA parameters were optimized using the trial and error method such as the population size. The obtained results indicated that the optimized ANN with two hidden layers and two neurons in each hidden layer can predict the resilient modulus of fiber-reinforced HMA with high accuracy (correlation coefficient: .96).

Enhancing the Predicting Accuracy of the Water Stage Using a Physical-Based Model and an Artificial Neural Network-Genetic Algorithm in a River System

Accurate simulations of river stages during typhoon events are critically important for flood control and are necessary for disaster prevention and water resources management in Taiwan. This study applies two artificial neural network (ANN) models, including the back propagation neural network (BPNN) and genetic algorithm neural network (GANN) techniques, to improve predictions from a one-dimensional flood routing hydrodynamic model regarding the water stages during typhoon events in the Danshuei River system in northern Taiwan. The hydrodynamic model is driven by freshwater discharges at the upstream boundary conditions and by the water levels at the downstream boundary condition. The model provides a sound physical basis for simulating water stages along the river. The simulated results of the hydrodynamic model show that the model cannot reproduce the water stages at different stations during typhoon events for the model calibration and verification phases. The BPNN and GANN models can improve the simulated water stages compared with the performance of the hydrodynamic model. The GANN model satisfactorily predicts water stages during the training and verification phases and exhibits the lowest values of mean absolute error, root-mean-square error and peak error compared with the simulated results at different stations using the hydrodynamic model and the BPNN model. Comparison of the simulated results shows that the GANN model can be successfully applied to predict the water stages of the Danshuei River system during typhoon events.

Adaptive GA-NN for MDF Prediction Model

This research presents a hybrid Genetic Algorithm Neural Network (GA-NN) model to replace the physical tests procedures of Medium Density Fiberboard (MDF). Data included in the model is MDF properties and its fiber characteristics. Multilayer Perceptron (MLP) NN model is reliable to learn from seven inputs fed to the network to produce prediction of three targets. In order to avoid result from local optimum scenario, GA optimizes synaptic weights of the network towards reducing prediction error. The research used a fixed probability rates for crossover and mutation for hybrid GA-NN model. GA-NN model is further improved using adaptive mechanism to help identify the best probability rates. The fitness value refers to Sum of Squared Error. Performance comparisons are among three models; namely NN with Back Propagation (BP), hybrid GA-NN and hybrid GA-NN with adaptive mechanism. Results show the hybrid GA-NN model perform much better than NN model used with back propagation optimizer. Adaptive mechanism in GA helps increase capability to converge at zero sooner than the ordinary GA.

Yarn engineering using hybrid artificial neural network-genetic algorithm model

This work aims to manufacture cotton yarns with requisite quality by choice of suitable raw materials for a given spinning system. To fulfill this aim, a hybrid model based on Artificial Neural Network (ANN) and Genetic Algorithm (GA) has been developed which captures both the high prediction power of ANN and global solution searching ability of GA. In an attempt to achieve a yarn having predefined tenacity and evenness, a constrained optimization problem is formulated with the ANN input-output relation between fibre and yarn properties. GA has been used to solve the optimization problem by searching the best combination of fibre properties that can translate into reality a yarn with the desired quality. The model is capable in identifying the set of fibre properties that gives requisite yarn quality with reasonable degree of accuracy.

Preparation of Carvedilol Nanoparticles by Emulsification Method and Optimization of Drug Release: Surface Response Design Versus Genetic Algorithm

The objectives of this research were the production of Eudragit nanoparticles of carvedilol, an anti-hypertension drug, for enhancement of its absorption and optimization of drug release. Nanoparticles were prepared by emulsification-solvent evaporation or diffusion methods. The statistical surface response design, based on the Box-Behnken model, was applied to evaluate the effect of four variables, each in two levels, on specifications of nanoparticles. An intelligent modeling system was established according to genetic algorithm to predict drug release from the nanoparticles. The neural network-genetic algorithm model showed a more precise method than surface response design in the prediction of the release properties of carvedilol from Eudragit nanoparticles.

MODELING AND OPTIMIZATION OF NUTRITIONALLY ENRICHED SOY WHEY GENERATION

ABSTRACTThe present investigation deals with the modeling and optimization of acid curdling of soymilk for soy whey preparation focusing on maximization of protein and polyphenol content as well as yield of soy whey. The findings, in this first report on fruit acid (citric acid and tartaric acid)‐based curdling, reveal a better alternative to the salt (calcium sulphate) curdling, extensively practiced in industries, in terms of nutritional enrichment. Reasonably accurate neural network models (regression coefficient ranging from 0.96 to 0.97) using citric acid and tartaric acids as mild fruit acids and trichloroacetic acid as standard protein coagulant was obtained based on the experimental data. The optimized set of process conditions was predicted through genetic algorithm and the effectiveness of the artificial neural network genetic algorithm model, validated through experiments, was indicated by significant correlations (regression coefficient ranging from 0.89 to 0.92) and nominal mean square error (0.000071–0.01).PRACTICAL APPLICATIONSThe work aims at devising a methodology for fruit acid‐based curdling of soymilk for nutritionally enriched soy whey generation. This opens up a new dimension for valorized utilization of soy whey, a potent pollutant, for formulation of functional foods as it is an excellent source of proteins with balanced amino acid profile and polyphenols. Moreover organic acid curdled soy whey, owing to its oligosaccharide content, can be utilized as a medium for probiotic fermentation for development of synbiotic supplements. The fruit acids used (citric acid and tartaric acid) can be obtained from common fruits abundantlyavailable, and hence, can economize the curdling process in industrial settings simultaneously adding to the nutritional profile.

Hybrid genetic algorithm-neural network: Feature extraction for unpreprocessed microarray data

Suitable techniques for microarray analysis have been widely researched, particularly for the study of marker genes expressed to a specific type of cancer. Most of the machine learning methods that have been applied to significant gene selection focus on the classification ability rather than the selection ability of the method. These methods also require the microarray data to be preprocessed before analysis takes place. The objective of this study is to develop a hybrid genetic algorithm-neural network (GANN) model that emphasises feature selection and can operate on unpreprocessed microarray data. The GANN is a hybrid model where the fitness value of the genetic algorithm (GA) is based upon the number of samples correctly labelled by a standard feedforward artificial neural network (ANN). The model is evaluated by using two benchmark microarray datasets with different array platforms and differing number of classes (a 2-class oligonucleotide microarray data for acute leukaemia and a 4-class complementary DNA (cDNA) microarray dataset for SRBCTs (small round blue cell tumours)). The underlying concept of the GANN algorithm is to select highly informative genes by co-evolving both the GA fitness function and the ANN weights at the same time. The novel GANN selected approximately 50% of the same genes as the original studies. This may indicate that these common genes are more biologically significant than other genes in the datasets. The remaining 50% of the significant genes identified were used to build predictive models and for both datasets, the models based on the set of genes extracted by the GANN method produced more accurate results. The results also suggest that the GANN method not only can detect genes that are exclusively associated with a single cancer type but can also explore the genes that are differentially expressed in multiple cancer types. The results show that the GANN model has successfully extracted statistically significant genes from the unpreprocessed microarray data as well as extracting known biologically significant genes. We also show that assessing the biological significance of genes based on classification accuracy may be misleading and though the GANN's set of extra genes prove to be more statistically significant than those selected by other methods, a biological assessment of these genes is highly recommended to confirm their functionality.