Accuracy Of Artificial Neural Network Research Articles

Accurate Prediction of Type II Diabetes using Artificial Neutral Networks

Diabetes is a serious and progressive condition that is rapidly increasing in incidence and currently ranks third on the list of all causes of mortality throughout the globe. A key challenge for any nation, but particularly for one that is undergoing substantial change is the high diabetes prevalence rate. Research in the field of epidemiology has demonstrated that obesity and Type II diabetes are the result of a combination of genetic predisposition and lifestyle factors such as bad eating habits and a lack of physical exercise. This article presents machine learning and feature selection enabled framework for diabetes type 2 prediction. This article uses artificial neural network for classification and prediction of diabetes type 2 data. Input data used in experiment is gathered from Pima Indian Diabetes Dataset. Results are compared on the basis of certain parameters like- accuracy, sensitivity, specificity. Accuracy of artificial neural network is better for classification and prediction of type 2 diabetes.

Predicting Dengue Outbreak based on Meteorological Data Using Artificial Neural Network and Decision Tree Models

Dengue fever is well-known as a potentially fatal disease, and the number of cases in some areas remains uncontrolled. Despite efforts to prevent the dengue outbreak from spreading further, vectors may be to blame. Identifying what weather characteristics contribute to dengue outbreaks is important to predict the dengue outbreak. This study proposes Artificial Neural Network (ANN) and Decision Tree (DT) models based on maximum temperature, minimum temperature, total rainfall, and average humidity to predict the dengue outbreak in Kota Bharu. Different numbers of hidden nodes were used in ANN to optimize the model. Both models, ANN and DT are evaluated based on accuracy, sensitivity and specificity showing that ANN (Accuracy = 68.85%, Sensitivity = 99.71%, Specificity = 1.27%), performed better than DT (Accuracy = 67.46%, Sensitivity = 98.82%, Specificity = 2.53%). This means that ANN outperforms DT when predicting a dengue outbreak in Kota Bharu. Based on the ANN model, it can be concluded that the number of hidden nodes affects the model's accuracy. Selecting the ideal number of hidden nodes for modeling the ANN model is appropriate. Even though ANN accuracy for prediction models is greater than DT, it is still low. It can be inferred that selecting a prediction model appropriate for a variety of dataset types and levels of complexity is important. Based on these models, the government may take pre-emptive actions to enhance public awareness about climate change.

A novel methodology for neural compact modeling based on knowledge transfer

This work presents a novel approach of using knowledge transfer to increase the accuracy of artificial neural network (ANN)-based device compact models, or neural compact models. This is useful when the amount of data available for training an ANN is limited. By utilizing relatively abundant data of a previous technology node, physical phenomena that are not evident in the limited data of the target technology node (e.g. gate-induced drain leakage) are accurately predicted. When meta learning algorithms are used, the accuracy of the model significantly increases, with relative linear error 10 times lower compared to the case when prior knowledge is not incorporated. The proposed methodology can be used to model future generation devices with limited data, utilizing data from well-characterized past technology node devices.

Enhancing Height Predictions of Brazilian Pine for Mixed, Uneven-Aged Forests Using Artificial Neural Networks

Artificial intelligence (AI) seeks to simulate the human ability to reason, make decisions, and solve problems. Several AI methodologies have been introduced in forestry to reduce costs and increase accuracy in estimates. We evaluate the performance of Artificial Neural Networks (ANN) in estimating the heights of Araucaria angustifolia (Bertol.) Kuntze (Brazilian pine) trees. The trees are growing in Uneven-aged Mixed Forests (UMF) in southern Brazil and are under different levels of competition. The dataset was divided into training and validation sets. Multi-layer Perceptron (MLP) networks were trained under different Data Normalization (DN) procedures, Neurons in the Hidden Layer (NHL), and Activation Functions (AF). The continuous input variables were diameter at breast height (DBH) and height at the base of the crown (HCB). As a categorical input variable, we consider the sociological position of the trees (dominant–SP1 = 1; codominant–SP2 = 2; and dominated–SP3 = 3), and the continuous output variable was the height (h). In the hidden layer, the number of neurons varied from 3 to 9. Results show that there is no influence of DN in the ANN accuracy. However, the increase in NHL above a certain level caused the model’s over-fitting. In this regard, around 6 neurons stood out, combined with logistic sigmoid AF in the intermediate layer and identity AF in the output layer. Considering the best selected network, the following values of statistical criteria were obtained for the training dataset (R2 = 0.84; RMSE = 1.36 m, and MAPE = 6.29) and for the validation dataset (R2 = 0.80; RMSE = 1.49 m, and MAPE = 6.53). The possibility of using categorical and numerical variables in the same modeling has been motivating the use of AI techniques in different forestry applications. The ANN presented generalization and consistency regarding biological realism. Therefore, we recommend caution when determining DN, amount of NHL, and using AF during modeling. We argue that such techniques show great potential for forest management procedures and are suggested in other similar environments.

Trainability and Accuracy of Artificial Neural Networks: An Interacting Particle System Approach

AbstractNeural networks, a central tool in machine learning, have demonstrated remarkable, high fidelity performance on image recognition and classification tasks. These successes evince an ability to accurately represent high‐dimensional functions, but rigorous results about the approximation error of neural networks after training are few. Here we establish conditions for global convergence of the standard optimization algorithm used in machine learning applications, stochastic gradient descent (SGD), and quantifying the scaling of its error with the size of the network. This is done by reinterpreting SGD as the evolution of a particle system with interactions governed by a potential related to the objective or “loss” function used to train the network. We show that, when the number n of units is large, the empirical distribution of the particles descends on a convex landscape towards the global minimum at a rate independent of n, with a resulting approximation error that universally scales as O(n−1). These properties are established in the form of a law of large numbers and a central limit theorem for the empirical distribution. Our analysis also quantifies the scale and nature of the noise introduced by SGD and provides guidelines for the step size and batch size to use when training a neural network. We illustrate our findings on examples in which we train neural networks to learn the energy function of the continuous 3‐spin model on the sphere. The approximation error scales as our analysis predicts in as high a dimension as d = 25. © 2022 Courant Institute of Mathematics and Wiley Periodicals LLC.

Acquisition and Automated Segmentation of Inertia Sensor Data for Mobile Camptocormia Assessment.

The camptocormia angle has been established as a strong indicator for evaluating the progress of Parkinson's disease and the efficacy of therapeutical approaches. A wearable setup is proposed to measure the camptocormia angle with the perpendicular method using five inertial sensors. This study identifies suitable inertial measurement unit sensors for mobile long-term measurement. Moreover, a machine-learning approach is presented for segmenting the recorded data into periods with different dominant activities. An artificial neural network was the better classifier compared to a support vector machine to recognize certain common activities in patients with camptocormia. The artificial neural network's accuracy, sensitivity, and F1-score were 92.4 %, 82.9 %, and 82.1 %, respectively. Clinical Relevance- The presented approach is expected to lead to a wearable system for long-term monitoring of the progress of camptocormia, yielding improved parameters compared to the conventional static photo method.

ANN modelling of surface roughness of FDM parts considering the effect of hidden layers, neurons, and process parameters

ABSTRACT Hidden layers perform a vital role in the performance of artificial neural networks (ANNs), especially for serpentine problems where the curtailment of accuracy and time complexity is observed. A higher number of hidden layers and neurons enlarge the order of weights and influence the ANN accuracy. In the present work, an ANN model is developed to predict the surface roughness of fused deposition modelling (FDM) parts considering the effect of the number of neurons, hidden layers, layer height, infill density, nozzle temperature, and print speed. A feedforward back-propagation machine learning algorithm is used to model an ANN. This study finds better prediction accuracy with the ANN architecture having two hidden layers with 150 neurons on each layer (R-squared: 0.875) followed by one hidden layer with 250 neurons (R-squared: 0.83). This study finds a decrease in the surface roughness of FDM parts with the increase in infill density. However, the surface roughness was observed to increase with the layer height, print speed, and nozzle temperature. This study finds a decrease in the surface roughness of FDM parts with the increase in infill density and an increase in the surface roughness with the layer height, print speed, and nozzle temperature.

Comfort evaluation of ZnO coated fabrics by artificial neural network assisted with golden eagle optimizer model

This paper introduces a novel technique to evaluate comfort properties of zinc oxide nanoparticles (ZnO NPs) coated woven fabrics. The proposed technique combines artificial neural network (ANN) and golden eagle optimizer (GEO) to ameliorate the training process of ANN. Neural networks are state-of-the-art machine learning models used for optimal state prediction of complex problems. Recent studies showed that the use of metaheuristic algorithms improve the prediction accuracy of ANN. GEO is the most advanced methaheurstic algorithm inspired by golden eagles and their intelligence for hunting by tuning their speed according to spiral trajectory. From application point of view, this study is a very first attempt where GEO is applied along with ANN to improve the training process of ANN for any textiles and composites application. Furthermore, the proposed algorithm ANN with GEO (ANN-GEO) was applied to map out the complex input-output conditions for optimal results. Coated amount of ZnO NPs, fabric mass and fabric thickness were selected as input variables and comfort properties were evaluated as output results. The obtained results reveal that ANN-GEO model provides high performance accuracy than standard ANN model, ANN models trained with latest metaheuristic algorithms including particle swarm optimizer and crow search optimizer, and conventional multiple linear regression.

Machine learning methods to predict particulate matter PM 2.5.

Introduction Pollution of air in urban cities across the world has been steadily increasing in recent years. An increasing trend in particulate matter, PM 2.5, is a threat because it can lead to uncontrollable consequences like worsening of asthma and cardiovascular disease. The metric used to measure air quality is the air pollutant index (API). In Malaysia, machine learning (ML) techniques for PM 2.5 have received less attention as the concentration is on predicting other air pollutants. To fill the research gap, this study focuses on correctly predicting PM 2.5 concentrations in the smart cities of Malaysia by comparing supervised ML techniques, which helps to mitigate its adverse effects. Methods In this paper, ML models for forecasting PM 2.5 concentrations were investigated on Malaysian air quality data sets from 2017 to 2018. The dataset was preprocessed by data cleaning and a normalization process. Next, it was reduced into an informative dataset with location and time factors in the feature extraction process. The dataset was fed into three supervised ML classifiers, which include random forest (RF), artificial neural network (ANN) and long short-term memory (LSTM). Finally, their output was evaluated using the confusion matrix and compared to identify the best model for the accurate prediction of PM 2.5. Results Overall, the experimental result shows an accuracy of 97.7% was obtained by the RF model in comparison with the accuracy of ANN (61.14%) and LSTM (61.77%) in predicting PM 2.5. Discussion RF performed well when compared with ANN and LSTM for the given data with minimum features. RF was able to reach good accuracy as the model learns from the random samples by using decision tree with the maximum vote on the predictions.

Deep Learning Based Centerline-Aggregated Aortic Hemodynamics: An Efficient Alternative to Numerical Modeling of Hemodynamics

Image-based patient-specific modelling of hemodynamics are gaining increased popularity as a diagnosis and outcome prediction solution for a variety of cardiovascular diseases. While their potential to improve diagnostic capabilities and thereby clinical outcome is widely recognized, these methods require considerable computational resources since they are mostly based on conventional numerical methods such as computational fluid dynamics (CFD). As an alternative to the numerical methods, we propose a machine learning (ML) based approach to calculate patient-specific hemodynamic parameters. Compared to CFD based methods, our approach holds the benefit of being able to calculate a patient-specific hemodynamic outcome instantly with little need for computational power. In this proof-of-concept study, we present a deep artificial neural network (ANN) capable of computing hemodynamics for patients with aortic coarctation in a centerline aggregated (i.e., locally averaged) form. Considering the complex relation between vessels shape and hemodynamics on the one hand and the limited availability of suitable clinical data on the other, a sufficient accuracy of the ANN may however not be achieved with available data only. Another key aspect of this study is therefore the successful augmentation of available clinical data. Using a statistical shape model, additional training data was generated which substantially increased the ANN's accuracy, showcasing the ability of ML based methods to perform in-silico modelling tasks previously requiring resource intensive CFD simulations.

Machine learning-based modeling of the coupling effect of strain rate and temperature on strain hardening for 5182-O aluminum alloy

This research characterizes the dynamic hardening behavior of an aluminum alloy sheet of 5182-O for the coupling effect of strain rate and temperature. Tests are carried out for dogbone specimens at different loading conditions to experimentally characterize the strain rate hardening and thermal softening effect for the alloy. The behaviours are then modeled by the Johnson-Cook, Zerilli-Armstrong and Lim-Huh models. In addition, the FEA-friendly polynomial model and artificial neural network (ANN) model are used to describe the highly non-linearity and coupling of strain hardening. Factors affecting ANN predicting accuracy and numerical computing efficiency are comprehensively studied including network structure, parameter settings and optimization algorithms. All the analytical and ANN models are also implemented into ABAQUS/Explicit to numerically compute the reaction force of tensile tests of dogbone specimens. The strain hardening curves are predicted by the analytical and ANN models for the comparison with experimental measurements to evaluate their performance. The experimental results show that the strain rate is slightly negative at room temperature, while the strain rate effect turns to be positive as temperature rises. The comparison of the flow curves between prediction and experiments reveals that the coupling effect is reasonably illustrated by the proposed polynomial model and the ANN model illustrates the flow curves with the dramatically much better accuracy than all the other models. The numerically predicted reaction forces prove that the ANN model accurately illustrates the load capability with the best agreement among the models studied in this research. The numerical computation also shows that the numerical computation efficiency of the ANN model is slightly reduced compared with analytical models, but the reduction is not much and worthwhile compared with the high accuracy of ANN.

Forecasting Inflation, Exchange Rate, and GDP using ANN and ARIMA Models: Evidence from Pakistan

Purpose: The purpose of this study is to specify an efficient forecast model for the accurate prediction of macroeconomic variables in the context of Pakistan.
 Design/Methodology/Approach: We particularly investigate the comparative accuracy of Artificial Neural Network (ANN) and Autoregressive Integrated Moving Average (ARIMA) models-based predictions using monthly data of inflation, exchange rate, and GDP from 1990 to 2014.
 Findings: According to our findings, the ANN-based forecasted inflation series is more precise as compared to ARIMA-based estimates. On the contrary, the ARIMA model outperforms the ANN model for exchange rate forecasts with the forecasted values being very close to the actual values. Further, ARIMA performs comparatively better in forecasting GDP with relatively smaller forecast error. On the whole, our findings suggest the ARIMA model provides appropriate results for forecasting exchange rates and GDP, while the ANN model offers precise estimates of inflation.
 Implications/Originality/Value: Our findings have important implications for the analysts and policymakers highlighting the need to use appropriate forecasting models that are well aligned with the structure of an economy.

New recommendation to predict export value using big data and machine learning technique

Official statistics on monthly export values have a publicity lag between the current period and the published publication. None of the previous researchers estimated the value of exports for the monthly period. This circumstance is due to limitations in obtaining supporting data that can predict the criteria for the current export value of goods. AIS data is one type of big data that can provide solutions in producing the latest indicators to forecast export values. Statistical Methods and Conventional Machine Learning are implemented as forecasting methods. Seasonal ARIMA and Artificial Neural Network (ANN) methods are both used in research to forecast the value of Indonesia’s exports. However, ANN has a weakness that requires high computational costs to obtain optimal parameters. Genetic Algorithm (GA) is effective in increasing ANN accuracy. Based on these backgrounds, this paper aims to develop and select an AIS indicator to predict the monthly export value in Indonesia and optimize ANN performance by combining the ANN algorithm with the genetic algorithm (GA-ANN). The research successfully established five indicators that can be used as predictors in the forecasting model. According to the model evaluation results, the genetic algorithm has succeeded in improving the performance of the ANN model as indicated by the resulting RMSE GA-ANN value, which is smaller than the RMSE of the ANN model.

Evaluation of Classification for Project Features with Machine Learning Algorithms

Due to the asymmetry of project features, it is difficult for project managers to make a reliable prediction of the decision-making process. Big data research can establish more predictions through the results of accurate classification. Machine learning (ML) has been widely applied for big data analytic and processing, which includes model symmetry/asymmetry of various prediction problems. The purpose of this study is to achieve symmetry in the developed decision-making solution based on the optimal classification results. Defects are important metrics of construction management performance. Accordingly, the use of suitable algorithms to comprehend the characteristics of these defects and train and test massive data on defects can conduct the effectual classification of project features. This research used 499 defective classes and related features from the Public Works Bid Management System (PWBMS). In this article, ML algorithms, such as support vector machine (SVM), artificial neural network (ANN), decision tree (DT), and Bayesian network (BN), were employed to predict the relationship between three target variables (engineering level, project cost, and construction progress) and defects. To formulate and subsequently cross-validate an optimal classification model, 1015 projects were considered in this work. Assessment indicators showed that the accuracy of ANN for classifying the engineering level is 93.20%, and the accuracy values of SVM for classifying the project cost and construction progress are 85.32% and 79.01%, respectively. In general, the SVM yielded better classification results from these project features. This research was based on an ML algorithm evaluation system for buildings as a classification model for project features with the goal of aiding project managers to comprehend defects.

A General Framework for Building Surrogate Models for Uncertainty Quantification in Computational Electromagnetics

In uncertainty analysis, surrogate modeling techniques demonstrate high efficiency and reliable precision in estimating the uncertainty for the finite difference time domain (FDTD) computation. However, building an accurate surrogate model may require a considerable number of system simulations which could be computationally expensive. To reduce such computational cost to build an accurate model, a general framework to build surrogate models for the FDTD computation in the human body based on the least angle regression (LARS) method and the artificial neural network (ANN) is proposed. The LARS method is adapted to dynamically select a number of informative random parameters which are significantly relevant to system outputs. We design a series of convergence criteria for ANN and introduce the adaptive moment estimation (ADAM) optimizer to ANN in order to improve the computational efficiency and accuracy of ANN. This is the first dynamic surrogate modeling technique for the FDTD computation designed by taking both accuracy and computational cost into account.

Estimation of hypersonic vehicle weight using Physics-Informed neural network supported by knowledge based engineering

Design space exploration is usually required during the conceptual design of hypersonic vehicle. Artificial neural network (ANN) is widely used for issues of design space exploration. However, gaps such as the low prediction accuracy of ANN with small training data and inefficiency of data generation are still unaddressed. In this paper, for investigation of the variation of the maximum takeoff weight (MTOW) of hypersonic vehicles as the configuration changes, a method combing physics-informed neural network (PINN) and knowledge based engineering (KBE) is developed to study the issue. In this method, PINN is used as the predictor of MTOW, the prior knowledge of engineers is able to be embedded into the loss function and guide the training of PINN. Compared with that of traditional neural network, the prediction accuracy of PINN is higher especially with small training data. Moreover, the KBE is used to guide the design process of generating training data and the efficiency of data generation can be improved. In order to account for the validity of the proposed method, the MTOW prediction of hypersonic vehicles with different wing configurations has been investigated. The results showed that the prediction accuracy of neural network can be improved with small training data.

Engineering synaptic plasticity through the control of oxygen vacancy concentration for the improvement of learning accuracy in a Ta2O5 memristor

Ta2O5 memristors exhibit bipolar switching properties attributable to the growth and destruction of oxygen vacancy filaments (OVFs). The transmission properties of biological synapse are mimicked in these memristors. The Ta2O5 memristor that contains numerous oxygen vacancies (OVs) is heated under N2 at 10 Torr, and it shows high conductance modulation linearity (CML) because the variation of OVF is governed by the redox reaction. The recognition accuracy of artificial neural networks (ANNs) is affected significantly by the CML of the memristor. Simulation using a convolutional neural network reveals that this Ta2O5 memristor exhibits a high learning accuracy of 93% because of its high CML. Spike-timing-dependent plasticity (STDP) was realized in Ta2O5 memristors. The change rate of synaptic weight variation in the STDP curve, which is also related to the learning accuracy of ANNs, is large in the Ta2O5 memristor heated under N2 at 10 Torr; this confirms that this memristor has a good learning accuracy. Spike rate-dependent plasticity and the transition from short-term plasticity to long-term plasticity are observed in Ta2O5 memristors. Further, they were obtained at a small potentiation spike in a Ta2O5 memristor heated under N2 at 10 Torr because numerous OVs exist in this memristor.

A comparative analysis of artificial neural network and support vector machine for online transient stability prediction considering uncertainties

ABSTRACT Power system transient stability is an integral part of power system planning and operation. Conventional approaches to assess transient stability are time consuming and hence, are not suitable for online application. Thus, this paper presents a comparative analysis of two different machine learning (ML) algorithms, i.e. artificial neural network (ANN) and support vector machine (SVM), for online transient stability prediction, considering various uncertainties (load, network topology, fault type, fault location, and fault clearing time). Time-domain simulations were conducted using DIgSILENT PowerFactory software for obtaining the training data for ML algorithms. MATLAB was used to apply the ML algorithms (ANN and SVM), and to draw a comparison between them. The classification accuracy of ANN and SVM was found to be 98.4% and 94.4% (fixed topology), respectively. The classification accuracy of ANN and SVM was found to be 97.7% and 93.2% (topology change), respectively. The training time of ANN was less than that of SVM (for both topology cases). These results for the IEEE 14-bus system demonstrated that both ANN and SVM can rapidly estimate the transient stability, considering uncertainties, with a reasonable accuracy; however, ANN outperformed SVM as its classification performance and computational performance was determined to be superior.

Combined Neural Network and Fuzzy Control for Maximum Power Point Tracking In Solar PV System

The most effective renewable energy in the future among energy. Using photovoltaic (PV) panels is the most effective approach to utilize solar energy for electrical power. The photovoltaic panel that uses solar energy has non-linear voltage-current and voltage-control properties. The control yield from the solar-powered PV board is highest at a certain voltage point. The maximum power point voltage is the voltage at which the PV board produces the most power. The construction of the solar-powered PV board and a description of its features have been presented. In order to acquire the highest control point voltage, the difference between the real Maximum Power Point Tracking (MPPT) voltage and the MPPT voltage deduced from the Artificial Neural Network (ANN) arrangement is displayed, and it is successfully accepted. The result appears to be an exact match for the accuracy of ANN. The PV board's highest control point can be successfully and accurately tracked using the ANN display that was obtained. From the various results obtained, it becomes obvious that the suggested computation proves to be considerably simpler in following the PV board's most extreme control point. The reaction time is drastically reduced when the proposed control approach is used since the PV voltage closely tracks the highest control point voltage. Additionally, the proposed strategy's precision is incredibly logical. In every irradiance and temperature scenario, the control framework performs well.

Improved artificial neural networks for 3D body posture and lumbosacral moment predictions during manual material handling activities

Body posture measurement approaches, required in biomechanical models to assess risk of musculoskeletal injuries, are usually costly and/or impractical for use in real workplaces. Therefore, we recently developed three artificial neural networks (ANNs), based on measured posture data on several individuals, to predict whole body 3D posture (coordinates of 15 markers located on body’s main joints), segmental orientations (Euler angles of 14 body segments), and lumbosacral (L5-S1) moments during static manual material handling (MMH) activities (ANNPosture, ANNAngle, and ANNMoment, respectively). These ANNs require worker's body height, body weight (only for ANNMoment), hand-load 3D position, and its mass as inputs to accurately predict 3D marker coordinates (RMSE = 7.0 cm), segmental orientations (RMSE = 29.9°) and L5-S1 moments (RMSE = 16.5 N.m) for various static MMH activities. The current work aims to further improve the accuracy of these ANNs by performing outlier elimination and data normalization (as effective tools to improve the accuracy of ANNs) as well as by introducing participant’s knee flexion angle (i.e., lifting technique: stoop, semi-squat, and full-squat) and body weight as new inputs into these ANNs. Results indicate that the RMSE of the new ANNPosture, ANNAngle, and ANNMoment reduced by, respectively, ∼43%, 10%, and 29% (from 7.0 cm, 29.9°, and 16.5 Nm in the original ANNs to, respectively, 4.0 cm, 27.0°, and 11.8 Nm). Such significant improvements in the predictive power of our ANNs further confirm their effectiveness as alternative posture-prediction approaches requiring minimal in vivo data collection in real workplaces.