Search results

Diabetic retinopathy (DR) is a central root of blindness all over the world. Though DR is tough to diagnose in starting stages, and the detection procedure might be time-consuming even for qualified experts. Nowadays, intelligent disease detection techniques are extremely acceptable for progress analysis and recognition of various diseases. Therefore, a computer-aided diagnosis scheme based on intelligent learning approaches is intended to propose for diagnosing DR effectively using a benchmark dataset.

The proposed DR diagnostic procedure involves four main steps: (1) image pre-processing, (2) blood vessel segmentation, (3) feature extraction, and (4) classification. Initially, the retinal fundus image is taken for pre-processing with the help of Contrast Limited Adaptive Histogram Equalization (CLAHE) and average filter. In the next step, the blood vessel segmentation is carried out using a segmentation process with optimized gray-level thresholding. Once the blood vessels are extracted, feature extraction is done, using Local Binary Pattern (LBP), Texture Energy Measurement (TEM based on Laws of Texture Energy), and two entropy computations – Shanon's entropy, and Kapur's entropy. These collected features are subjected to a classifier called Neural Network (NN) with an optimized training algorithm. Both the gray-level thresholding and NN is enhanced by the Modified Levy Updated-Dragonfly Algorithm (MLU-DA), which operates to maximize the segmentation accuracy and to reduce the error difference between the predicted and actual outcomes of the NN. Finally, this classification error can correctly prove the efficiency of the proposed DR detection model.

The overall accuracy of the proposed MLU-DA was 16.6% superior to conventional classifiers, and the precision of the developed MLU-DA was 22% better than LM-NN, 16.6% better than PSO-NN, GWO-NN, and DA-NN. Finally, it is concluded that the implemented MLU-DA outperformed state-of-the-art algorithms in detecting DR.

This paper adopts the latest optimization algorithm called MLU-DA-Neural Network with optimal gray-level thresholding for detecting diabetic retinopathy disease. This is the first work utilizes MLU-DA-based Neural Network for computer-aided Diabetic Retinopathy diagnosis.

The purpose of this paper is to build a neural network system to predict corporate credit rating in Jordanian non-financial firms, using 19 different financial characteristics such as profitability, leverage ratios, liquidity, bankruptcy, and sales performance.

The study adopts two neural network techniques namely, Kohonen network and Back Propagation Neural Network (BPNN). Our sample includes the manufacturing firms that have provided the required financial information for the period from 2000 to 2007.

BPNN has successfully predicted firms with high performance gaining A rating and the bankrupted firms with D rating for the period from 2005 to 2007.

This study is the first study to investigate credit rating in Jordan using Neural Network technique.

Energy management is critical to data centres (DCs) majorly because they are high energy-consuming facilities and demand for their services continue to rise due to rapidly increasing global demand for cloud services and other technological services. This projected sectoral growth is expected to translate into increased energy demand from the sector, which is already considered a major energy consumer unless innovative steps are used to drive effective energy management systems. The purpose of this study is to provide insights into the expected energy demand of the DC and the impact each measured parameter has on the building's energy demand profile. This serves as a basis for the design of an effective energy management system.

This study proposes novel tunicate swarm algorithm (TSA) for training an artificial neural network model used for predicting the energy demand of a DC. The objective is to find the optimal weights and biases of the model while avoiding commonly faced challenges when using the backpropagation algorithm. The model implementation is based on historical energy consumption data of an anonymous DC operator in Cape Town, South Africa. The data set provided consists of variables such as ambient temperature, ambient relative humidity, chiller output temperature and computer room air conditioning air supply temperature, which serve as inputs to the neural network that is designed to predict the DC’s hourly energy consumption for July 2020. Upon preprocessing of the data set, total sample number for each represented variable was 464. The 80:20 splitting ratio was used to divide the data set into training and testing set respectively, making 452 samples for the training set and 112 samples for the testing set. A weights-based approach has also been used to analyze the relative impact of the model’s input parameters on the DC’s energy demand pattern.

The performance of the proposed model has been compared with those of neural network models trained using state of the art algorithms such as moth flame optimization, whale optimization algorithm and ant lion optimizer. From analysis, it was found that the proposed TSA outperformed the other methods in training the model based on their mean squared error, root mean squared error, mean absolute error, mean absolute percentage error and prediction accuracy. Analyzing the relative percentage contribution of the model's input parameters based on the weights of the neural network also shows that the ambient temperature of the DC has the highest impact on the building’s energy demand pattern.

The proposed novel model can be applied to solving other complex engineering problems such as regression and classification. The methodology for optimizing the multi-layered perceptron neural network can also be further applied to other forms of neural networks for improved performance.

Based on the forecasted energy demand of the DC and an understanding of how the input parameters impact the building's energy demand pattern, neural networks can be deployed to optimize the cooling systems of the DC for reduced energy cost.

The use of TSA for optimizing the weights and biases of a neural network is a novel study. The application context of this study which is DCs is quite untapped in the literature, leaving many gaps for further research. The proposed prediction model can be further applied to other regression tasks and classification tasks. Another contribution of this study is the analysis of the neural network's input parameters, which provides insight into the level to which each parameter influences the DC’s energy demand profile.

The purpose of this paper is to examine the feasibility of committee neural network (CNN) theory for the improvement of accuracy and consistency of the neural network model on the estimation of preconsolidation pressure from the field piezocone measurements.

In this study, CNN theory is introduced to improve the initial weight dependency of the neural network model on the prediction of preconsolidation pressure of soft clay from a piezocone test result. It was found that the proposed CNN model can improve the initial weight dependency of the NN model and provide a more consistent and precise inference result than existing NN models, as well as empirical and theoretical models.

It was found that the CNN overcomes the initial weight dependency of the single neural network model. Various committees of the single multilayer perceptrons (MLPs) were tested. It was found that if eight single MLPs, which have the same structure but have been trained with a different initial weight and bias, are accumulated in the committee with the same weighting factor, any variation on the prediction of the preconsolidation pressure from the piezocone test result can be simply and successfully eliminated.

In recent years, ANN has been found to be a powerful theory for analyzing complex relationships involving a multitude of variables, on many geotechnical applications. However, single MLP, when repeatedly trained on the same patterns, tends to reach different minima of the objective function each time and hence give a different set of neuron weights, because the solution is not unique for noisy data, as in most geotechnical problems. The authors observed that a committee neural network system is able to provide improved performance compared with a single optimal neural network.

Mechanical products usually require deterministic finite element analysis in the design phase to determine whether their structures meet the requirements. However, deterministic design ignores the influence of uncertainties in the design and manufacturing process of mechanical products, leading to the problem of a lack of design safety or excessive redundancy in the design. In order to improve the accuracy and rationality of the design results, a robust design method for structural reliability based on an active-learning marine predator algorithm (MPA)–backpropagation (BP) neural network is proposed.

The MPA was used to obtain the optimal weights and thresholds of a BP neural network, and an active-learning function applicable to neural networks was proposed to efficiently improve the prediction performance of the BP neural network. On this basis, a robust optimization design method for mechanical product reliability based on the active-learning MPA-BP model was proposed. Random moving quadrilateral sampling was used to obtain the sample points required for training and testing of the neural network, and the reliability sensitivity corresponding to each sample point was calculated by subset simulated significant sampling (SSIS). The total mass of the mechanical product and the structural reliability sensitivity of the trained active-learning MPA-BP model output were taken as the optimization objectives, and a multi-objective reliability-robust optimization design model was constructed, which was solved by the second-generation non-dominated ranking genetic algorithm (NSGA-II). Then, the dominance function was used in the obtained Pareto solution set to make a dominance-seeking decision to obtain the final reliability-robust optimization design solution. The feasibility of the proposed method was verified by a reliability-robust optimization design example of the bogie frame.

The prediction error of the active-learning MPA-BP neural network was smaller than those of the particle swarm optimization (PSO)-BP, marine predator algorithm (MPA)-BP and genetic algorithm (GA)-BP neural networks under the same basic parameter settings of the algorithm, which indicated that the improvement strategy proposed in this paper improved the prediction accuracy of the BP neural network. To ensure the reliability of the bogie frame, the reliability sensitivity and total mass of the bogie frame were reduced, which not only realized the lightweight design of the bogie frame, but also improved the reliability and robustness of the bogie.

The MPA algorithm with a higher optimization efficiency was introduced to find the weights and thresholds of the BP neural network. A new active-learning function was proposed to improve the prediction accuracy of the MPA-BP neural network.

The purpose of this paper is to devise a new approach to synthesize closed‐loop feedback guidance law for online thrust‐insensitive optimal trajectory generation utilizing neural networks.

The proposed methodology utilizes an open‐loop variational formulation that initially determines optimal launch/ascent trajectories for various scenarios of known uncertainties in the thrust profile of typical solid propellant engines. These open‐loop optimized trajectories will then provide the knowledge base needed for the subsequent training of a neural network. The trained network could eventually produce thrust‐insensitive closed‐loop optimal guidance laws and trajectories in flight.

The proposed neuro‐optimal guidance scheme is effective for online closed‐loop optimal path planning through some measurable and computable engine and flight parameters.

Determination of closed‐loop optimal guidance law for non‐linear dynamic systems with uncertainties in system and environment has been a challenge for researchers and engineers for many years. The problem of steering a solid propellant driven vehicle is one of these challenges. Even though a few researchers have worked in the area of non‐linear optimal control and thrust‐insensitive guidance, this paper proposes a new strategy for the determination of closed‐loop online thrust insensitive guidance laws leading to optimal flight trajectories for solid propellant launch and ascent vehicles.

Recently, more and more attention has been put forth on the application and deep learning, due to the widespread practicability of neural network computation. The purpose of this paper is developing an effective algorithm to automatically discover the optimal neural network architecture for several real applications.

The author proposes a novel algorithm, namely, progressive genetic-based neural architecture search (PG-NAS), as a solution to efficiently find the optimal neural network structure for given data. PG-NAS also employs several operations to effectively shrink the search space to reduce the computation cost and improve the accuracy validation.

The proposed PG-NAS could be utilized on several tasks for discovering the optimal network structure. The author reduces the demand of manual settings when implementing artificial intelligence (AI) models; hence, PG-NAS requires less human intervention than traditional machine learning. The average and top-1 metrics, such as error, loss and accuracy, are used to measure the discovered neural architectures of the proposed model over all baselines. The experimental results show that, with several real datasets, the proposed PG-NAS model consistently outperforms the state-of-the-art models in all metrics.

Generally, the size and the complexity of the search space for the neural network dominates the performance of computation time and resources. In this study, PG-NAS utilizes genetic operations to effectively generate the compact candidate set, i.e. fewer combinations need to be generated when constructing the candidate set. Moreover, by the proposed selector in PG-NAS, the non-promising network structure could be significantly pruned off. In addition, the accuracy derivation of each combination in the candidate set is also a performance bottleneck. The author develops a predictor network to efficiently estimate the accuracy to avoid the time-consuming derivation. The learning of the prediction process is also adjusted dynamically; this adaptive learning of the predictor could capture the pattern of training data effectively and efficiently. Furthermore, the proposed PG-NAS algorithm is applied on several real datasets to show its practicability and scalability.

The purpose of this paper is three-fold: to review the categories explaining mainly optimization algorithms (techniques) in that needed to improve the generalization performance and learning speed of the Feedforward Neural Network (FNN); to discover the change in research trends by analyzing all six categories (i.e. gradient learning algorithms for network training, gradient free learning algorithms, optimization algorithms for learning rate, bias and variance (underfitting and overfitting) minimization algorithms, constructive topology neural networks, metaheuristic search algorithms) collectively; and recommend new research directions for researchers and facilitate users to understand algorithms real-world applications in solving complex management, engineering and health sciences problems.

The FNN has gained much attention from researchers to make a more informed decision in the last few decades. The literature survey is focused on the learning algorithms and the optimization techniques proposed in the last three decades. This paper (Part II) is an extension of Part I. For the sake of simplicity, the paper entitled “Machine learning facilitated business intelligence (Part I): Neural networks learning algorithms and applications” is referred to as Part I. To make the study consistent with Part I, the approach and survey methodology in this paper are kept similar to those in Part I.

Combining the work performed in Part I, the authors studied a total of 80 articles through popular keywords searching. The FNN learning algorithms and optimization techniques identified in the selected literature are classified into six categories based on their problem identification, mathematical model, technical reasoning and proposed solution. Previously, in Part I, the two categories focusing on the learning algorithms (i.e. gradient learning algorithms for network training, gradient free learning algorithms) are reviewed with their real-world applications in management, engineering, and health sciences. Therefore, in the current paper, Part II, the remaining four categories, exploring optimization techniques (i.e. optimization algorithms for learning rate, bias and variance (underfitting and overfitting) minimization algorithms, constructive topology neural networks, metaheuristic search algorithms) are studied in detail. The algorithm explanation is made enriched by discussing their technical merits, limitations, and applications in their respective categories. Finally, the authors recommend future new research directions which can contribute to strengthening the literature.

The FNN contributions are rapidly increasing because of its ability to make reliably informed decisions. Like learning algorithms, reviewed in Part I, the focus is to enrich the comprehensive study by reviewing remaining categories focusing on the optimization techniques. However, future efforts may be needed to incorporate other algorithms into identified six categories or suggest new category to continuously monitor the shift in the research trends.

The authors studied the shift in research trend for three decades by collectively analyzing the learning algorithms and optimization techniques with their applications. This may help researchers to identify future research gaps to improve the generalization performance and learning speed, and user to understand the applications areas of the FNN. For instance, research contribution in FNN in the last three decades has changed from complex gradient-based algorithms to gradient free algorithms, trial and error hidden units fixed topology approach to cascade topology, hyperparameters initial guess to analytically calculation and converging algorithms at a global minimum rather than the local minimum.

The existing literature surveys include comparative study of the algorithms, identifying algorithms application areas and focusing on specific techniques in that it may not be able to identify algorithms categories, a shift in research trends over time, application area frequently analyzed, common research gaps and collective future directions. Part I and II attempts to overcome the existing literature surveys limitations by classifying articles into six categories covering a wide range of algorithm proposed to improve the FNN generalization performance and convergence rate. The classification of algorithms into six categories helps to analyze the shift in research trend which makes the classification scheme significant and innovative.

The measurement capabilities of the data envelopment analysis (DEA) models are used to train the artificial neural network (ANN) models for the best performance modeling of the sawmills in Ontario. The bootstrap DEA models measure robust technical efficiency scores and have benchmarking abilities, whereas the ANN models use abstract learning from a limited set of information and provide the predictive power.

The complementary modeling approaches of the DEA and the ANN provide an adaptive decision support tool for each sawmill.

The trained ANN models demonstrate promising results in predicting the relative efficiency scores and the optimal combination of the inputs and the outputs for three categories (large, medium and small) of sawmills in Ontario. The average absolute error in predicting the relative efficiency scores varies from 0.01 to 0.04, and the predicted optimal combination of the inputs (roundwood and employees) and the output (lumber) demonstrate that a large percentage of the sawmills shows less than 10% error in the prediction results.

The purpose of this study is to develop an integrated DEA-ANN model that can help in the continuous improvement and performance evaluations of the forest industry working under uncertain business environment.

To present a neural network‐based approach to the design of electromagnetic devices.

A neural model is created which reproduces the relationship between the design parameters of the device and the performance parameters, typically field values.

The neural model is a single hidden layer MLP network, trained by using a set of cases calculated, for example, by means of a finite element analysis. The design problem can be solved by fixing the performance values at the output of the network and by calculating the corresponding input values. The relationship between the input and the output of the neural network is represented by three equations systems. By means of these three systems, we can forward the domain of the input, and we can back propagate the desired output throughout the network layers. In such a way, both the domain of the design parameters and the domain of the desired performances values can be projected in the same space. Whatever point inside the intersection between the two projected domains corresponds to a solution of the design problem.

Presents a procedure which is able to find a point belonging to such an intersection.