Abstract Background To classify chest computed tomography (CT) images as positive or negative for coronavirus disease 2019 (COVID-19) quickly and accurately, researchers attempted to develop effective models by using medical images. Results A convolutional neural network (CNN) ensemble model was developed for classifying chest CT images as positive or negative for COVID-19. To classify chest CT images acquired from COVID-19 patients, the proposed COVID19-CNN ensemble model combines the use of multiple trained CNN models with a majority voting strategy. The CNN models were trained to classify chest CT images by transfer learning from well-known pre-trained CNN models and by applying their algorithm hyperparameters as appropriate. The combination of algorithm hyperparameters for a pre-trained CNN model was determined by uniform experimental design. The chest CT images (405 from COVID-19 patients and 397 from healthy patients) used for training and performance testing of the COVID19-CNN ensemble model were obtained from an earlier study by Hu in 2020. Experiments showed that, the COVID19-CNN ensemble model achieved 96.7% accuracy in classifying CT images as COVID-19 positive or negative, which was superior to the accuracies obtained by the individual trained CNN models. Other performance measures (i.e., precision, recall, specificity, and F1-score) obtained bythe COVID19-CNN ensemble model were higher than those obtained by individual trained CNN models. Conclusions The COVID19-CNN ensemble model had superior accuracy and excellent capability in classifying chest CT images as COVID-19 positive or negative.

Abstract Background Researchers have attempted to apply deep learning methods of artificial intelligence for rapidly and accurately detecting acute lymphoblastic leukemia (ALL) in microscopic images. Results A Resnet101-9 ensemble model was developed for classifying ALL in microscopic images. The proposed Resnet101-9 ensemble model combined the use of the nine trained Resnet-101 models with a majority voting strategy. Each trained Resnet-101 model integrated the well-known pre-trained Resnet-101 model and its algorithm hyperparameters by using transfer learning method to classify ALL in microscopic images. The best combination of algorithm hyperparameters for the pre-trained Resnet-101 model was determined by Taguchi experimental method. The microscopic images used for training of the pre-trained Resnet-101 model and for performance tests of the trained Resnet-101 model were obtained from the C-NMC dataset. In experimental tests of performance, the Resnet101-9 ensemble model achieved an accuracy of 85.11% and an F 1 -score of 88.94 in classifying ALL in microscopic images. The accuracy of the Resnet101-9 ensemble model was superior to that of the nine trained Resnet-101 individual models. All other performance measures (i.e., precision, recall, and specificity) for the Resnet101-9 ensemble model exceeded those for the nine trained Resnet-101 individual models. Conclusion Compared to the nine trained Resnet-101 individual models, the Resnet101-9 ensemble model had superior accuracy in classifying ALL in microscopic images obtained from the C-NMC dataset.

BACKGROUND: To diagnose key pathologies of age-related macular degeneration (AMD) and diabetic macular edema (DME) quickly and accurately, researchers attempted to develop effective artificial intelligence methods by using medical images. RESULTS: A convolutional neural network (CNN) with transfer learning capability is proposed and appropriate hyperparameters are selected for classifying optical coherence tomography (OCT) images of AMD and DME. To perform transfer learning, a pre-trained CNN model is used as the starting point for a new CNN model for solving related problems. The hyperparameters (parameters that have set values before the learning process begins) in this study were algorithm hyperparameters that affect learning speed and quality. During training, different CNN-based models require different algorithm hyperparameters (e.g., optimizer, learning rate, and mini-batch size). Experiments showed that, after transfer learning, the CNN models (8-layer Alexnet, 22-layer Googlenet, 16-layer VGG, 19-layer VGG, 18-layer Resnet, 50-layer Resnet, and a 101-layer Resnet) successfully classified OCT images of AMD and DME. CONCLUSIONS: The experimental results further showed that, after transfer learning, the VGG19, Resnet101, and Resnet50 models with appropriate algorithm hyperparameters had excellent capability and performance in classifying OCT images of AMD and DME.

BACKGROUND: Researchers have tried to identify and count different blood cells in microscopic smear images by using deep learning methods of artificial intelligence to solve the highly time-consuming problem. RESULTS: The three types of blood cells are platelets, red blood cells, and white blood cells. This study used the Resnet50 network as a backbone network of the single shot detector (SSD) for automatically identifying and counting different blood cells and, meanwhile, proposed a systematic method to find a better combination of algorithm hyperparameters of the Resnet50 network for promoting accuracy for identifying and counting blood cells. The Resnet50 backbone network of the SSD with its optimized algorithm hyperparameters, which is called the Resnet50-SSD model, was developed to enhance the feature extraction ability for identifying and counting blood cells. Furthermore, the algorithm hyperparameters of Resnet50 backbone networks of the SSD were optimized by the Taguchi experimental method for promoting detection accuracy of the Resnet50-SSD model. The experimental result shows that the detection accuracy of the Resnet50-SSD model with 512 × 512 × 3 input images was better than that of the Resnet50-SSD model with 300 × 300 × 3 input images on the test set of blood cells images. Additionally, the detection accuracy of the Resnet50-SSD model using the combination of algorithm hyperparameters got by the Taguchi method was better than that of the Resnet50-SSD model using the combination of algorithm hyperparameters given by the Matlab example. CONCLUSION: In blood cell images acquired from the BCCD dataset, the proposed Resnet50-SSD model had higher accuracy in identifying and counting blood cells, especially white blood cells and red blood cells.

BACKGROUND: The prevalence of chronic disease is growing in aging societies, and artificial-intelligence-assisted interpretation of macular degeneration images is a topic that merits research. This study proposes a residual neural network (ResNet) model constructed using uniform design. The ResNet model is an artificial intelligence model that classifies macular degeneration images and can assist medical professionals in related tests and classification tasks, enhance confidence in making diagnoses, and reassure patients. However, the various hyperparameters in a ResNet lead to the problem of hyperparameter optimization in the model. This study employed uniform design-a systematic, scientific experimental design-to optimize the hyperparameters of the ResNet and establish a ResNet with optimal robustness. RESULTS: An open dataset of macular degeneration images ( https://data.mendeley.com/datasets/rscbjbr9sj/3 ) was divided into training, validation, and test datasets. According to accuracy, false negative rate, and signal-to-noise ratio, this study used uniform design to determine the optimal combination of ResNet hyperparameters. The ResNet model was tested and the results compared with results obtained in a previous study using the same dataset. The ResNet model achieved higher optimal accuracy (0.9907), higher mean accuracy (0.9848), and a lower mean false negative rate (0.015) than did the model previously reported. The optimal ResNet hyperparameter combination identified using the uniform design method exhibited excellent performance. CONCLUSION: The high stability of the ResNet model established using uniform design is attributable to the study's strict focus on achieving both high accuracy and low standard deviation. This study optimized the hyperparameters of the ResNet model by using uniform design because the design features uniform distribution of experimental points and facilitates effective determination of the representative parameter combination, reducing the time required for parameter design and fulfilling the requirements of a systematic parameter design process.

BACKGROUND: Heart sound measurement is crucial for analyzing and diagnosing patients with heart diseases. This study employed phonocardiogram signals as the input signal for heart disease analysis due to the accessibility of the respective method. This study referenced preprocessing techniques proposed by other researchers for the conversion of phonocardiogram signals into characteristic images composed using frequency subband. Image recognition was then conducted through the use of convolutional neural networks (CNNs), in order to classify the predicted of phonocardiogram signals as normal or abnormal. However, CNN requires the tuning of multiple hyperparameters, which entails an optimization problem for the hyperparameters in the model. To maximize CNN robustness, the uniform experiment design method and a science-based methodical experiment design were used to optimize CNN hyperparameters in this study. RESULTS: An artificial intelligence prediction model was constructed using CNN, and the uniform experiment design method was proposed to acquire hyperparameters for optimal CNN robustness. The results indicate Filters ([Formula: see text]), Stride ([Formula: see text]), Activation functions ([Formula: see text]), and Dropout ([Formula: see text]) to be significant factors considerably influencing the ability of CNN to distinguish among heart sound states. Finally, the confirmation experiment was conducted, and the hyperparameter combination for optimal model robustness was Filters ([Formula: see text]) = 32, Kernel Size ([Formula: see text] = 3 × 3, Stride ([Formula: see text]) = (1,1), Padding ([Formula: see text] as same, Optimizer ([Formula: see text] as the stochastic gradient descent, Activation functions ([Formula: see text]) as relu, and Dropout ([Formula: see text]) = 0.544. With this combination of parameters, the model had an average prediction accuracy rate of 0.787 and standard deviation of 0. CONCLUSION: In this study, phonocardiogram signals were used for the early prediction of heart diseases. The science-based and methodical uniform experiment design was used for the optimization of CNN hyperparameters to construct a CNN with optimal robustness. The results revealed that the constructed model exhibited robustness and an acceptable accuracy rate. Other literature has failed to address hyperparameter optimization problems in CNN; a method is subsequently proposed for robust CNN optimization, thereby solving this problem.

BACKGROUND: Dengue epidemics is affected by vector-human interactive dynamics. Infectious disease prevention and control emphasize the timing intervention at the right diffusion phase. In such a way, control measures can be cost-effective, and epidemic incidents can be controlled before devastated consequence occurs. However, timing relations between a measurable signal and the onset of the pandemic are complex to be discovered, and the typical lag period regression is difficult to capture in these complex relations. This study investigates the dynamic diffusion pattern of the disease in terms of a probability distribution. We estimate the parameters of an epidemic compartment model with the cross-infection of patients and mosquitoes in various infection cycles. We comprehensively study the incorporated meteorological and mosquito factors that may affect the epidemic of dengue fever to predict dengue fever epidemics. RESULTS: We develop a dual-parameter estimation algorithm for a composite model of the partial differential equations for vector-susceptible-infectious-recovered with exogeneity compartment model, Markov chain Montel Carlo method, and boundary element method to evaluate the epidemic periodicity under the effect of environmental factors of dengue fever, given the time series data of 2000-2016 from three cities with a population of 4.7 million. The established computer model of "energy accumulation-delayed diffusion-epidemics" is proven to be effective to predict the future trend of reported and unreported infected incidents. Our artificial intelligent algorithm can inform the authority to cease the larvae at the highest vector infection time. We find that the estimated dengue report rate is about 20%, which is close to the number of official announcements, and the percentage of infected vectors increases exponentially yearly. We suggest that the executive authorities should seriously consider the accumulated effect among infected populations. This established epidemic prediction model of dengue fever can be used to simulate and evaluate the best time to prevent and control dengue fever. CONCLUSIONS: Given our developed model, government epidemic prevention teams can apply this platform before they physically carry out the prevention work. The optimal suggestions from these models can be promptly accommodated when real-time data have been continuously corrected from clinics and related agents.

Nanofluid/ultrasonic atomization minimal quantity lubrication (MQL) was used in an SKD11 steel micromilling process with ultrasonic dispersion. The nanofluid atomizer substantially improved nanofluid reunification and delivery to the cutting zone. The micromilling cutting force and temperature were used to calculate the signal-to-noise ratio (S/N). Normalization of the S/N in the gray relational analysis obtained a gray correlation coefficient, which is used in a fuzzy analysis to obtain a multiple performance characteristic indexes (MPCI). The MPCI is maximized by optimizing the process parameters for multiple quality characteristics and then comparing the optimized parameters with those in the MQL system. Experimental results indicate that the nanofluid parameters containing the density of nanofluid, feed rate, and distance of nozzle are the most influential control factors, while 3.617 N and 117.1 °C, respectively, are the best combination of parameters. Different lubricating methods are also compared in terms of effects on micromilling cutting force, micromilling temperature, micromilling tool wear, and surface burr. Experiments show that nanofluid multiwall carbon nanotube/ultrasonic atomization MQL obtained the best results.

An adaptive network-based fuzzy inference system (ANFIS) with a sliding-level particle swarm optimization (SL-PSO) is proposed for optimizing parameters of a chemicalmechanical process for polishing a color filter (CMP-CF). The SL-PSO is used not only to find the best membership function types, but also to optimize the premise and consequent parameters for ANFIS. The important process parameters for CMP-CF included the initial time, polishing time, polishing pad weight (down force), slurry chemicals, and rotation speed. The output targets were red pixels, green pixels, blue pixels, and the surface roughness. First, the performance of the SL-PSO was tested with 18 continuous global numerical optimization problems, including six unimodal functions, seven multimodal functions, and five complex rotated and shifted functions. Nonparametric Wilcoxon tests were also used in multiple-problem analysis for simultaneous comparison of various algorithms over a problem set. The computational experiments showed that the proposed SL-PSO approach outperforms PSO-based methods reported in the literature. Finally, the proposed SL-PSO method was used to optimize CMP-CF parameters. The experimental results showed that the ANFIS with SL-PSO outperforms the conventional ANFIS method and conventional back propagation neural network in terms of prediction accuracy. A practical industrial application in a CF manufacturer showed that the ANFIS with SL-PSO obtained superior results compared to the previous method and immediately enhanced production efficiency. Together, these experimental results indicate that the proposed ANFIS with SL-PSO is a reliable method for optimizing CMP-CF processes.

This paper presents a new intelligent data-driven adaptive method (IDAM) of performing automatic online searches in real time. An online implementation of the proposed IDAM achieved rapid real-time optimization of system integration scaling factors for an automatic touch panel lamination machine. The proposed IDAM combines three-level orthogonal arrays (OAs), signal-to-noise ratios (SNRs), the best combined strategy, and a stepwise ratio. Three-level OA experiments with factor values are used to perform positional experiments, and SNRs are calculated for each experimental value. After the best combination of factor values (in terms of factor effect) is determined, new three-level factor values are derived by applying a stepwise ratio and used in further three-level OA experiments. These steps are repeated until the stopping criterion is met. Compared to conventional methods, the use of the IDAM in practical industrial applications, i.e., online real-time precision positioning for automatic touch panel lamination machines, reduces the number of experiments needed to obtain the system integration scaling factors that minimize the iteration count. For example, the IDAM required less than 40 online real-time experiments with a specific stepwise ratio for system integration scaling factors that met the minimum requirement of two iterations. In 50 independent experimental runs using the robust scaling factors obtained by the method, an average of 2.15 iterations was needed to achieve a positional accuracy within 5 μm. The main advantage of the proposed IDAM over conventional methods is its effectiveness for automatically finding robust parameters for online alignment systems in real time and with fewer experiments.

This paper used an improved particle swarm optimizer to solve an problem for optimal multiobjective PID design, including the uncertainty of the controlled field and external unknown interference. Experimental result shows that the performance of optimal design for the PID controller by using the improved PSO is better than by using the hybrid Taguchi-genetic algorithm.

The purpose of this paper is to make a nanofluid/ultrasonic atomization minimum quality lubrication (MQL) system, and to explore the effectiveness of the system used in the grinding process of SKD11 mold steel with heat treatment. Besides, a set of optimized multiple quality characteristics of grinding parameter combinations is obtained by using the Taguchi method and the fuzzy inference system. The experimental parameter combinations design is performed by using L18(21× 37) orthogonal table of the Taguchi method, where the experimental parameters are the type of nanoparticles, nanofluid concentration, tangential velocity, table rate, nozzle angle, nozzle distance, air pressure, and spray volume. The single-quality characteristics such as grinding force ratio, grinding temperature, and surface roughness are considered in this paper. The gray relation analysis and the fuzzy inference system are adopted to obtain the multiple performance characteristic index (MPCI), then the combination of maximum MPCI is the optimized multiple quality characteristics of grinding parameter combinations. This paper also compares the difference of nanofluid/ultrasonic atomization MQL and nanofluid/air MQL through the optimized multiple quality characteristics of grinding parameter combinations. Experimental results indicate that the nanofluid parameters containing the type of nanoparticles, nanofluid concentration, and spray volume are the most influential control factors. Comparing nanofluid/ultrasonic atomization MQL with nanofluid/air MQL can also get the best of grinding force ratio, grinding temperature, surface roughness, and surface morphology.

This study rewrote a fractional-order particle swarm optimizer algorithmic equation and used an improved uniform design method (IUDM) to find the best combination for parameters of FPSO. Compared to PSO, FPSO makes a high convergence rate. In the improved FPSO, there are 4 parameters to influence effectiveness. Uniform design is an experimental method and suitable for multiple parameters and multiple level experiments. IUDM is improved from uniform design method as per using the best combination strategy and a stepwise ratio. And then, IUDM is used to find the best combination for parameter values of FPSO. In the experiments, some test functions are used to find the best combination for parameter values by using IUDM and also verify the effectiveness of the improved FPSO. From the results, the best combination for parameter values of FPSO can be obtain by IUDM and the proposed FPSO obtain the better performance than PSO and original FPSO according the best combination.

A soft-computing technology is proposed for optimizing the positional compensation parameters for an exposure machine for flexible printed circuit boards (FPCBs). The proposed technology integrates a full-factorial experimental design, a multilayer perceptron (MLP) artificial neural network, and the Taguchi-based genetic algorithm (TBGA). First, a full-factorial experimental design is used to conduct experiments and to accumulate data that represent the positional compensation parameters of an exposure machine. The MLP is then used to build a positioning model of an exposure machine by minimizing the performance criterion of mean-squared error (mse). Finally, the TBGA is used to optimize the positional compensation parameters for the exposure machine. The experimental results demonstrate the excellent performance of the MLP-TBGA approach in obtaining positional compensation parameters for decreasing the number of iterations and the alignment time. For example, in 50 independent runs, the average number of iterations for precision positioning decreased from 4.5 to 3.2, and the alignment time decreased by 41%, if the required positional accuracy was 3 μm. In another experimental application for precision positioning in which the required positional accuracy was 5 μm, the average number of iterations required in 50 practical experiments decreased from 3.3 to 2.1, and the alignment time decreased by 57%. The main advantage of the proposed soft-computing approach is its potential use for solving related problems in widely varying industries.

A Taguchi sliding-based differential evolution algorithm with orthogonal array (TDEOA) is proposed for solving tolerance design problems. Tolerance affects system performance and leads to violation of design constraints. By including a Taguchi three-level orthogonal array, the proposed TDEOA obtains robust optimal solutions that minimize the impact of variations in machining variables and that maintain compliance with a comprehensive set of process constraints. After evaluating its performance in practical case studies of rough and finish grinding processes, the performance of the proposed TDEOA is compared with those of other nature-inspired optimization approaches. In addition, a distinct way has been introduced to estimate the reliability of a set of measurements for a surface grinding process. Reliability tests from the proposed TDEOA approach confirm its effectiveness as specified tolerances are considered.

This paper proposes a procedure for process parameters design by combining both modeling and optimization methods. The proposed procedure integrates the Taguchi method, the artificial neural network (ANN), and the genetic algorithm (GA). First, the Taguchi method is applied to minimize experimental numbers and to collect experimental data representing the quality performances of a system. Next, the ANN is used to build a system model based on the data from the Taguchi experimental method. Then, the GA is employed to search for the optimal process parameters. A process parameters design for a titanium dioxide (TiO2) thin film in the vacuum sputtering process is studied in this paper. The quality objective is to form a smaller water contact angle on the TiO2thin-film surface. The water contact angle is 4° obtained from the system model of the proposed procedure. The process parameters obtained from the proposed procedure were used to conduct the experiment in the vacuum sputtering process for the TiO2thin film. The water contact angle given from the practical experiment is 3.93°. The difference percent is 1.75% between 4° and 3.93°. The result obtained from the system model of the proposed procedure is promising. Hence, we can conclude that the proposed procedure is a very good approach in solving the problem of the process parameters design.

This paper considers the design problem of the robust quadratic-optimal parallel-distributed-compensation (PDC) controllers for Takagi–Sugeno (TS) fuzzy-model-based control systems with both elemental parametric uncertainties and norm-bounded approximation error. By complementarily fusing the robust stabilizability condition, the orthogonal functions approach (OFA), and the hybrid Taguchi genetic algorithm (HTGA), an integrative method is presented in this paper to design the robust quadratic-optimal PDC controllers such that 1) the uncertain TS-fuzzy-model-based control systems can be robustly stabilized, and 2) a quadratic integral performance index for the nominal TS-fuzzy-model-based control systems can be minimized. In this paper, the robust stabilizability condition is proposed in terms of linear matrix inequalities (LMIs). By using the OFA and the LMI-based robust stabilizability condition, the robust quadratic-optimal PDC control problem for the uncertain TS-fuzzy-model-based dynamic systems is transformed into a static constrained-optimization problem represented by the algebraic equations with constraint of LMI-based robust stabilizability condition, thus greatly simplifying the robust optimal PDC control design problem. Then, for the static constrained-optimization problem, the HTGA is employed to find the robust quadratic-optimal PDC controllers of the uncertain TS-fuzzy-model-based control systems. Two design examples of the robust quadratic-optimal PDC controllers for an uncertain inverted pendulum system and an uncertain nonlinear mass–spring–damper mechanical system are given to demonstrate the applicability of the proposed integrative approach.

This article presents a hybrid multiobjective genetic algorithm to aid the tracking of the daily aircraft schedule recovery problem under disturbance events such as severe weather and mechanical problems. The proposed algorithm extends from the original method of inequality-based multiobjective genetic algorithm (MMGA) and utilizes an adaptive evaluated vector (AEV) to co-work with MMGA efficiently when maintaining the Pareto set of recovered schedules in the evolutionary population. Two main goals would be presented: One is to provide a multi-objective solution to the recovery problem and the other is to address the performance requirement on the recovery approach. A simulated disturbance experiment on the practical aircraft schedule is made to validate the recovery results under the expected short-time period.

This paper considers the design problem of the robust quadratic-optimal parallel-distributed-compensation (PDC) controllers for the Takagi-Sugeno (TS) fuzzy-modelbased control systems with both elemental parametric uncertainties and norm-bounded approximation error. An integrative method, which complementarily fuses the robust stabilizability condition, the orthogonal-functions approach (OFA) and the hybrid Taguchi-genetic algorithm (HTGA), is presented in this paper to design the robust quadratic-optimal PDC controllers, in which the robust stabilizability condition is proposed in terms of linear matrix inequalities (LMIs).

In this paper, we propose a method of multi-objective optimization Evolutionary Computation by using Evaluated Preference Genetic Algorithm (EPGA). The method is applied to quickly solve a time-effective aircraft routing in response to the schedule disruption of short-haul flights and tried to optimize objective functions including flight connection, flight duty swap, total flight delay time, delayed flights and flights over 30 minutes delay. The proposed EPGA approach here is a novel genetic algorithm to effectively resolve multi-objective optimization problems; it can consider multiple objectives simultaneously and then explore the optimal solution. Traditionally, airline schedule disruption management problem is solved by Operations Research (OR) techniques, which always require a precise mathematical model. But in real-world airline operation environment, there are too many factors to be considered dynamically, and thus it is very difficult to define a precise mathematical model in time. In this research, we propose EPGA to deal with Robust Airline Schedule Recover Problem (RASRP) which is easier to model the practical problems. Furthermore, this method is verified by real flight schedules of Taiwan domestic airlines. The results show that the high quality solutions can be obtained in a few minutes. Therefore, EPGA can be used as a fast decision support tool for practical complex airline operations.

In this paper, we present an improved genetic algorithm to solve the worst-case circuit tolerance design problem, which has many design parameters and constraints. The evolutionary design approach, which is called a quality-engineering-based genetic algorithm (QEGA), with a penalty function is proposed for solving the constrained optimization problem. The QEGA approach is able to explore a wide design parameter space without any prior knowledge about position and size of the region of acceptability. The QEGA approach is a method of combining the traditional genetic algorithm (TGA), which has a powerful global exploration capability, with the quality engineering method, which can exploit the optimum offspring. The quality engineering method is inserted between crossover and mutation operations of the TGA. Then, the systematic reasoning ability of the quality engineering method is incorporated in the crossover operations to select the better genes to achieve crossover, and consequently enhance the genetic algorithms. Therefore, the QEGA approach can be more robust and quickly convergent. In the worst-case circuit tolerance design problem, the vertex analysis has been used to check the feasibility of any candidate tolerance region. For this reason, if, as for a large part of cases, the region of acceptability is convex and simply connected, the QEGA approach ensures an optimal design with 100% yield. The proposed QEGA approach with a penalty function is effectively applied to solve the worst-case circuit tolerance design problem. The computational experiments also show that the presented QEGA approach can obtain better results than previous design methods

In this paper, a hybrid Taguchi-genetic algorithm (HTGA) is applied to solve the problem of tuning both network structure and parameters of a feedforward neural network. The HTGA approach is a method of combining the traditional genetic algorithm (TGA), which has a powerful global exploration capability, with the Taguchi method, which can exploit the optimum offspring. The Taguchi method is inserted between crossover and mutation operations of a TGA. Then, the systematic reasoning ability of the Taguchi method is incorporated in the crossover operations to select the better genes to achieve crossover, and consequently enhance the genetic algorithms. Therefore, the HTGA approach can be more robust, statistically sound, and quickly convergent. First, the authors evaluate the performance of the presented HTGA approach by studying some global numerical optimization problems. Then, the presented HTGA approach is effectively applied to solve three examples on forecasting the sunspot numbers, tuning the associative memory, and solving the XOR problem. The numbers of hidden nodes and the links of the feedforward neural network are chosen by increasing them from small numbers until the learning performance is good enough. As a result, a partially connected feedforward neural network can be obtained after tuning. This implies that the cost of implementation of the neural network can be reduced. In these studied problems of tuning both network structure and parameters of a feedforward neural network, there are many parameters and numerous local optima so that these studied problems are challenging enough for evaluating the performances of any proposed GA-based approaches. The computational experiments show that the presented HTGA approach can obtain better results than the existing method reported recently in the literature.

A direct computational algorithm, by using the elegant operational properties of the shifted Chebyshev series (SCS), for solving the time-varying Takagi-Sugeno (TS) fuzzy-model-based feedback dynamic equations is first developed in this paper. The developed computational algorithm only involves the straightforward algebraic computation. Then, the developed algorithm is integrated with the hybrid Taguchi-genetic algorithm (HTGA) to design the optimal non-parallel-distributed-compensation (non-PDC) controller (linear state feedback controller) of the time-varying TS-fuzzy-model-based control systems under the criterion of minimizing a quadratic integral performance index, where the integral performance index is also converted into the algebraic form by using the SCS approach (SCSA). Thus, this proposed approach facilitates the design task of the non-PDC optimal controller for the time-varying TS-fuzzy-model-based control systems. A design example of the optimal non-PDC controller for the pendulum system with the vibration in the vertical direction on the pivot point is given to demonstrate the applicability of the proposed approach.

In this paper, a hybrid Taguchi-genetic algorithm (HTGA) is proposed to solve global numerical optimization problems with continuous variables. The HTGA combines the traditional genetic algorithm (TGA), which has a powerful global exploration capability, with the Taguchi method, which can exploit the optimum offspring. The Taguchi method is inserted between crossover and mutation operations of a TGA. Then, the systematic reasoning ability of the Taguchi method is incorporated in the crossover operations to select the better genes to achieve crossover, and consequently, enhance the genetic algorithm. Therefore, the HTGA can be more robust, statistically sound, and quickly convergent. The proposed HTGA is effectively applied to solve 15 benchmark problems of global optimization with 30 or 100 dimensions and very large numbers of local minima. The computational experiments show that the proposed HTGA not only can find optimal or close-to-optimal solutions but also can obtain both better and more robust results than the existing algorithm reported recently in the literature.