Variant Of Artificial Neural Network Research Articles

Remote Sensing and Machine Learning for Safer Railways: A Review

Regular railway inspections are crucial for maintaining their safety and efficiency. However, traditional inspection methods are complex and expensive. Consequently, there has been a significant shift toward combining remote sensing (RS) and machine learning (ML) techniques to enhance the efficiency and accuracy of railway defect monitoring while reducing costs. The advantages of RS-ML techniques include their ability to automate and refine inspection processes and address challenges such as image quality and methodological limitations. However, the integration of RS and ML in railway monitoring is an emerging field, with diverse methodologies and outcomes that the research has not yet synthesized. To fill this gap, this study conducted a systematic literature review (SLR) to consolidate the existing research on RS-ML applications in railway inspection. The SLR meticulously compiled and analyzed relevant studies, evaluating the evolution of research trends, methodological approaches, and the geographic distribution of contributions. The findings showed a notable increase in relevant research activity over the last five years, highlighting the growing interest in this realm. The key methodological patterns emphasize the predominance of approaches based on convolutional neural networks, a variant of artificial neural networks, in achieving high levels of precision. These findings serve as a foundational resource for academics, researchers, and practitioners in the fields of computer science, engineering, and transportation to help guide future research directions and foster the development of more efficient, accurate, and cost-effective railway inspection methods.

Comparing machine learning approaches for estimating soil saturated hydraulic conductivity.

Characterization of near (field) saturated hydraulic conductivity (Kfs) of the soil environment is among the crucial components of hydrological modeling frameworks. Since the associated laboratory/field experiments are time-consuming and labor-intensive, pedotransfer functions (PTFs) that rely on statistical predictors are usually integrated with the existing measurements to predict Kfs in other areas of the field. In this study some of the most appropriate machine learning approaches, including variants of artificial neural networks (ANNs) were used for predicting Kfs by some easily measurable soil attributes. The analyses were performed using 100 measurements in Bajgah Agricultural Experimental Station. First, physico-chemical inputs as bulk density (BD), initial water content (Wi), saturated water content (Ws), mean weight diameter (MWD), and geometric mean diameter (GMD) of aggregates, pH, electrical conductivity (EC), and calcium carbonate equivalent (CCE) were measured. Then, radial basis functions (RBFNNs), multilayer perceptron (MLPNNs), hybrid genetic algorithm (GA-NNs), and particle swarm optimization (PSO-NNs) neural networks were utilized to develop PTFs and compared their accuracy with the traditional regression model (MLR) using statistical indices. The statistical assessment indicated that PSO-NNs with the lowest RMSE and MAPE as well as the highest correlation coefficient (R) value provided the most accurate and robust prediction of Kfs. The prediction models ranked as PSO-NNs (R = 0.958; RMSE = 0.343; MAPE = 9.47), GA-NNs (R = 0.949; RMSE = 0.404; MAPE = 11.83), MLPNNs (R = 0.933; RMSE = 0.426; MAPE = 12.13), RBFNNs (R = 0.926; RMSE = 0.452; MAPE = 14.30), and MLR (R = 0.675; RMSE = 0.685; MAPE = 22.54) in terms of their performances for the test data set. Results revealed that all NN models particularly PSO-NNs were efficient in prediction of Kfs. However, further evaluations may be recommended for other soil conditions and input variables to quantify their potential uncertainties and wider potential and versatility before they are used in other geographical locations/soil conditions.

Solving partial differential equations with hybridized physic-informed neural network and optimization approach: Incorporating genetic algorithms and L-BFGS for improved accuracy

Partial differential equations (PDEs) are essential mathematical models for describing a wide range of physical phenomena. Numerically, Physic-Informed Neural Networks (PINNs), a variant of artificial neural networks, present a promising method for solving PDEs. However, due to limitation in accuracy and stability, various adaptive PINN variants have been proposed. We have designed a novel approach that adopted self-adaptive PINN (SA-PINN) with two optimization techniques: the genetic algorithm (GA) and the limited-memory Broyden-Fletcher-Goldfarb-Shanno (L-BFGS) algorithm. Self-adaptive PINN modifies the weights in the loss function to be fully trainable, enabling the ANN to learn and stabilize the PINN in approximating the difficult regions of the solution. GA initializes the population of ANN trainable parameters to optimize the training process with less number of iterations, while L-BFGS is used to find the best solution accurately. Our proposed approach, named SA-PINN-GA-LBFGS, is tested on solving several benchmark PDE problems including elliptic, parabolic, and hyperbolic types. We compare our results with state-of-the-art methods, demonstrating that SA-PINN-GA-LBFGS provides higher accuracy and greater efficiency.

Reconstruction Old Students Image Using The Autoencoder Method

Image Processing is image processing with a digital computer to produce new images according to the user's wishes. One implementation is to reconstruct the image. Through the extraction stages can get the characteristics of an image. The algorithm used is Adam Optimization, an extension of the stochastic gradient reduction that has seen wider adoption for deep learning applications in computer vision and natural language processing. In this study, we use the autoencoder technique, one variant of artificial neural networks generally used to "encode" data. The autoencoder is trained to produce the same output as the input. This image reconstruction aims to process an image whose quality is not very clear, to be precise. This, if possible, can be used to detect someone's face from photos. In reconstructing this image through the encode and decode process by defining Conv2D and Maxpool, it is processed into training with epoch 100 times while for the prediction process using Keras library. Then, the last one gets an accuracy of 0,022. The result is the output of the reconstructed image and calculation graph

A Smart Campus Internet-of-Things (IoT) Model for Smart Classroom Conditioning Using a Hybridized Technique

This research study presents Smart Campus (SC) Internet-of-Things (IoTs) enabled systems model that will support end-user and automatic functions for proper air conditioning of SC classrooms environment. It consists of a hybrid data learning predictor system using an emerging variant of Artificial Neural Network (ANN) called Neuronal Auditory Machine Intelligence (NeuroAMI) and a Linear Regressor (LR) of polynomial-order-of-1.The system was initially applied separately to the automated coordination of a smart bed in a laboratory sized classroom environment at a University Campus, and simulated using the high-level programming language – MATLAB, while end user interaction model was developed in the Java2ME programming language. Simulations results considering several trial runs showed that the ANN predictor generally performed better than the LR model with over 80% classification accuracy. While considering limited training data points, the LR predictor was found to be superior at one of the simulation trial runs. At 20% data point, the LR was activated while the NeuroAMI remains inactive, but above the 20% level, the NeuroAMI performed better. One advantage of this proposed hybrid system is the ability to deal with continuous data; exactly the same way human brains functions. This feat has not been possible in conventional ANN systems, especially in this area dealing with small data points.

Artificial Neural Network-Derived Unified Six-Dimensional Potential Energy Surface for Tetra Atomic Isomers of the Biogenic [H, C, N, O] System.

Recognition of different structural patterns in different potential energy surface regions, such as in isomerizing quasilinear tetra atomic molecules, is important for understanding the details of underlying physics and chemistry. In this respect, using three variants of artificial neural networks (ANNs), we investigated the six-dimensional (6-D) singlet potential energy surfaces (PES) of tetra atomic isomers of the biogenic [H, C, N, O] system. At first, we constructed a separate ANN potential for each of the studied isomers. In the next step, a comparative assessment of the separate ANN models led to the setting up of a unified 6-D singlet PES equally and accurately describing all studied isomers. The constructed unified model yields relative energies comparable to those obtained either from the gold standard CCSD(T) method or from separate ANNs for each of the studied isomers. The accuracy of the unified singlet PES is on the order of 10-4 Hartrees (0.1 kcal/mol). The developed PES in this work captures the main features of nonlinear and quasilinear tetra atomic isomers of this biogenic system.

Performance Evaluation of Various Solar Forecasting Models for Structural & Endogenous Datasets

The forecasting of solar irradiation with high precision is critical for fulfilling electricity demand. The dataset used to train the learning-based models has a direct impact on the model’s prediction accuracy. This work evaluates the impact of two types of datasets: structural and endogenous datasets over the prediction accuracy of different solar forecasting models (five variants of artificial neural network (ANN) based models, Support vector machine (SVM), Linear Regression, Bagged and Boosted Regression tree). The issue of variability estimation is also explored in the paper in order to choose the best model for a given dataset. The performance of the models is assessed using two essential error metrics: mean absolute percentage error (MAPE) and root mean square error (RMSE). The results shows that the MAPE and RMSE for structural data vary from 1.99% to 29.73% and 23.39 W/m2 to 165.21 W/m2, respectively, whereas these errors for endogenous dataset ranges from 1.98% to 31.19% and 23.64 W/m22 to 152.56 W/m22. Moreover, these findings, together with the data variability findings, suggest that SVM is the best model for all forms of data variability, whereas CFNN may be employed for greater variability.

Computational Intelligence Based Prediction of Alzheimer Disease from Imbalanced Mild Cognitive Impairment Samples

The disease of Alzheimer is a progressive neurological disease. In Alzheimer's disease, as neurons are injured and die throughout the brain. Dementia is the most important and primary concern for Alzheimer’s disease - there is deterioration in memory, thinking, behaviour and the ability to perform everyday activities. In the early stage of memory loss is the result of Mild cognitive impairment (MCI) and it also jeopardise the individual ability of performing his/her independently daily living activities. So, in later stage MCI patients may develop to Alzheimer disease (AD). In this paper a well-known Computational Intelligence tool called Artificial Neural Networks (ANNs) is used to predict AD from MCI samples. Two variants of ANNs - multilayer perceptron (MLP) and radial basis function (RBF) network are used for prediction. We have used 47 MCI samples (previously used by many researchers) for prediction problem. It has been observed that class distributions of MCI dataset are out of proportion i.e. lack of balance (imbalance). We have used oversampling algorithm with cross-validation for imbalanced MCI datasets to predict the developing of Alzheimer's or another dementia. In terms of prediction accuracy our findings are more relevant and it gives better results compared to previous studies which considered without imbalanced scenarios.

Reconstruction Old Students Image Using The Autoencoder Method

Image Processing is image processing with a digital computer to produce new images according to the user's wishes. One implementation is to reconstruct the image. Through the extraction stages are able to get the characteristics of an image. The algorithm used is Adam Optimization, which is an extension of the stochastic gradient reduction that has just seen wider adoption for deep learning applications in computer vision and natural language processing. In this study using the autoencoder technique, which is one variant of artificial neural networks that are generally used to "encode" data. Autoencoder is trained to be able to produce the same output as the input. This image reconstruction aims to process an image whose quality is not very clear to be clear. This if possible can be used to detect someone's face from a distance of photos. In reconstructing this image through the encode and decode process by defining Conv2D and Maxpool, it is processed into training with epoch 100 times while for the prediction process using Keras library. Then the last one gets an accuracy of 0,022. The final result is the output of the reconstructed image and calculation graph.

Multi-Tier Cellular Handover with Multi-Access Edge Computing and Deep Learning

This paper proposes a potential enhancement of handover for the next-generation multi-tier cellular network, utilizing two fifth-generation (5G) enabling technologies: multi-access edge computing (MEC) and machine learning (ML). MEC and ML techniques are the primary enablers for enhanced mobile broadband (eMBB) and ultra-reliable and low latency communication (URLLC). The subset of ML chosen for this research is deep learning (DL), as it is adept at learning long-term dependencies. A variant of artificial neural networks called a long short-term memory (LSTM) network is used in conjunction with a look-up table (LUT) as part of the proposed solution. Subsequently, edge computing virtualization methods are utilized to reduce handover latency and increase the overall throughput of the network. A realistic simulation of the proposed solution in a multi-tier 5G radio access network (RAN) showed a 40–60% improvement in overall throughput. Although the proposed scheme may increase the number of handovers, it is effective in reducing the handover failure (HOF) and ping-pong rates by 30% and 86%, respectively, compared to the current 3GPP scheme.

Novel hybrid model based on echo state neural network applied to the prediction of stock price return volatility

The prediction of stock price return volatilities is important for financial companies and investors to help to measure and managing market risk and to support financial decision-making. The literature points out alternative prediction models - such as the widely used heterogeneous autoregressive (HAR) specification - which attempt to forecast realized volatilities accurately. However, recent variants of artificial neural networks, such as the echo state network (ESN), which is a recurrent neural network based on the reservoir computing paradigm, have the potential for improving time series prediction. This paper proposes a novel hybrid model that combines HAR specification, the ESN, and the particle swarm optimization (PSO) metaheuristic, named HAR-PSO-ESN, which combines the feature design of the HAR model with the prediction power of ESN, and the consistent PSO metaheuristic approach for hyperparameters tuning. The proposed model is benchmarked against existing specifications, such as autoregressive integrated moving average (ARIMA), HAR, multilayer perceptron (MLP), and ESN, in forecasting daily realized volatilities of three Nasdaq (National Association of Securities Dealers Automated Quotations) stocks, considering 1-day, 5-days, and 21-days ahead forecasting horizons. The predictions are evaluated in terms of r-squared and mean squared error performance metrics, and the statistical comparison is made through a Friedman test followed by a post-hoc Nemenyi test. Results show that the proposed HAR-PSO-ESN hybrid model produces more accurate predictions on most of the cases, with an average R2 (coefficient of determination) of 0.635, 0.510, and 0.298, an average mean squared error of 5.78 × 10−8, 5.78 × 10−8, and 1.16 × 10−7, for 1, 5, and 21 days ahead on the test set, respectively. The improvement is statistically significant with an average rank of 1.44 considering the three different datasets and forecasting horizons.

Convolutional neural network acoustic model for robust Indonesian speech recognition in noisy environment

Noise causes the decreasing accuracy of automatic speech recognition (ASR). Several techniques have been developed and proposed to overcome this problem. Using artificial neural network (ANN) as acoustic model is one of the techniques. Convolutional neural network (CNN) is a variant of ANN that has been used for acoustic modeling. Another approach is to do pre-processing to the speech signal or to the extracted acoustic feature from speech signal, such as cepstral mean and variance normalization (CMVN). On this work, CNN acoustic models were trained by using CMVN pre-processed acoustic feature to make a noise-robust speech recognition system. Two group of models were made, each to handle 2 kinds of noise (babble noise and street noise). Those acoustic models were tested with noisy speech at different SNR (signal-to-noise ratio) value. Testing results from CNN acoustic models were compared with the ones from Gaussian Mixture Model-Hidden Markov Model (GMM-HMM) acoustic models. Testing results showed the increasing accuracy scores of acoustic models when models were trained using more variation of training data. CNN acoustic models that were trained using FBANK feature have higher accuracy scores than GMM-HMM models that were built using the same feature.

Prediction of compressive and flexural strengths of jarosite mixed cement concrete pavements using artificial neural networks

In this paper, an attempt has been made to apply and compare the prediction capability of two variants of Artificial Neural Networks: feed-forward neural network (FFNN) and the radial basis function network (RBFN) for modelling the flexural and compressive strengths of jarosite mixed cement concrete for pavements. The compressive strength and the flexural strength are dependent upon a total of eight inputs. Their values are experimentally determined from the specimens containing 0%, 10%, 15%, 20% and 25% of the jarosite as a partial replacement to cement in M40 concrete mix. These specimens had undergone a curing for 3, 7, 28, 90, 180 and 365 days providing the inputs–outputs experimental data for the 90 observations. The simulation experiments showed that the performance of FFNN is found to be better than that of the RBFN model as the former model provided lesser values of the performance indicators such as Mean Average Error and Mean Squared Error. Further, the FFNN model required a lesser number of parameters to be tuned during the training as compared to the RBFN model.

Tool condition monitoring in CNC end milling using wavelet neural network based on machine vision

The monitoring of tool condition in machining processes has significant importance to control the quality of machined parts and to reduce equipment downtime. This study investigates the application of a special variant of artificial neural networks (ANNs), in particular, wavelet neural network (WNN) for tool wear monitoring in CNC end milling process of high-speed steel. A mixed level design of experiments with machining parameters of cutting speed, feed rate, cutting depth, and machining time is developed, from which 126 experiments are conducted. For each experiment, tool wear and surface roughness of machined workpiece are measured. The tool wear images are processed, and the descriptor of wear zone is extracted. The WNN is then applied to predict the flank wear of the cutting tool and compared with commonly used types of ANNs and the statistical model. Different input combinations with the inclusion of wear zone descriptor and surface roughness of machined parts are used to evaluate the performance of all models. Results show that the WNN with the input parameters of cutting speed, feed rate, depth of cut, machining time, and descriptor of wear zone predicts the degree of tool wear most accurately.

Low‐complexity wireless sensor system for partial discharge localisation

This study describes a key element of any modern wireless sensor system: data processing. The authors describe a system consisting of a wireless sensor network and an algorithmic software for condition-based monitoring of electrical plant in a live substation. Specifically, the aim is to monitor for the presence of partial discharge (PD) using a matrix of inexpensive radio sensors with limited processing capability. A low-complexity fingerprinting technique is proposed, given that the sensor nodes to be deployed will be highly constrained in terms of processing power, memory and battery life. Two variants of artificial neural network (ANN) learning models (multilayer perceptron and generalised regression neural network) that use regression as a form of function approximation are developed and their performance compared to K-nearest neighbour and weighted K-nearest neighbour models. The results indicate that the ANN models yield superior performance in terms of robustness against noise and may be particularly suited for PD localisation.

Differentially Private Image Classification Using Support Vector Machine and Differential Privacy

The ubiquity of data, including multi-media data such as images, enables easy mining and analysis of such data. However, such an analysis might involve the use of sensitive data such as medical records (including radiological images) and financial records. Privacy-preserving machine learning is an approach that is aimed at the analysis of such data in such a way that privacy is not compromised. There are various privacy-preserving data analysis approaches such as k-anonymity, l-diversity, t-closeness and Differential Privacy (DP). Currently, DP is a golden standard of privacy-preserving data analysis due to its robustness against background knowledge attacks. In this paper, we report a scheme for privacy-preserving image classification using Support Vector Machine (SVM) and DP. SVM is chosen as a classification algorithm because unlike variants of artificial neural networks, it converges to a global optimum. SVM kernels used are linear and Radial Basis Function (RBF), while ϵ -differential privacy was the DP framework used. The proposed scheme achieved an accuracy of up to 98%. The results obtained underline the utility of using SVM and DP for privacy-preserving image classification.

Biogeography-based optimisation for data classification problems

Classification is a task of supervised learning whose aim is to identify to which of a set of categories a new input element belongs. Probabilistic neural network is a variant of artificial neural network, which is simple in structure, easy for training and often used in classification problems. In this paper, the authors propose an improved probabilistic neural network model that employs biogeography-based optimisation to enhance the accuracy of the classification. The proposed approach was tested on 11 standard benchmark medical datasets from the machine-learning repository. Results show that the classification accuracy of the proposed improved probabilistic neural network model outperforms that of the traditional probabilistic neural network model.

A Review of Short Term Load Forecasting using Artificial Neural Network Models

Abstract The electrical short term load forecasting has been emerged as one of the most essential field of research for efficient and reliable operation of power system in last few decades. It plays very significant role in the field of scheduling, contingency analysis, load flow analysis, planning and maintenance of power system. This paper addresses a review on recently published research work on different variants of artificial neural network in the field of short term load forecasting. In particular, the hybrid networks which is a combination of neural network with stochastic learning techniques such as genetic algorithm(GA), particle swarm optimization (PSO) etc. which has been successfully applied for short term load forecasting (STLF) is discussed thoroughly.

Quantification of sand fraction from seismic attributes using Neuro-Fuzzy approach

In this paper, we illustrate the modeling of a reservoir property (sand fraction) from seismic attributes namely seismic impedance, seismic amplitude, and instantaneous frequency using Neuro-Fuzzy (NF) approach. Input dataset includes 3D post-stacked seismic attributes and six well logs acquired from a hydrocarbon field located in the western coast of India. Presence of thin sand and shale layers in the basin area makes the modeling of reservoir characteristic a challenging task. Though seismic data is helpful in extrapolation of reservoir properties away from boreholes; yet, it could be challenging to delineate thin sand and shale reservoirs using seismic data due to its limited resolvability. Therefore, it is important to develop state-of-art intelligent methods for calibrating a nonlinear mapping between seismic data and target reservoir variables. Neural networks have shown its potential to model such nonlinear mappings; however, uncertainties associated with the model and datasets are still a concern. Hence, introduction of Fuzzy Logic (FL) is beneficial for handling these uncertainties. More specifically, hybrid variants of Artificial Neural Network (ANN) and fuzzy logic, i.e., NF methods, are capable for the modeling reservoir characteristics by integrating the explicit knowledge representation power of FL with the learning ability of neural networks. In this paper, we opt for ANN and three different categories of Adaptive Neuro-Fuzzy Inference System (ANFIS) based on clustering of the available datasets. A comparative analysis of these three different NF models (i.e., Sugeno-type fuzzy inference systems using a grid partition on the data (Model 1), using subtractive clustering (Model 2), and using Fuzzy c-means (FCM) clustering (Model 3)) and ANN suggests that Model 3 has outperformed its counterparts in terms of performance evaluators on the present dataset. Performance of the selected algorithms is evaluated in terms of correlation coefficients (CC), root mean square error (RMSE), absolute error mean (AEM) and scatter index (SI) between target and predicted sand fraction values. The achieved estimation accuracy may diverge minutely depending on geological characteristics of a particular study area. The documented results in this study demonstrate acceptable resemblance between target and predicted variables, and hence, encourage the application of integrated machine learning approaches such as Neuro-Fuzzy in reservoir characterization domain. Furthermore, visualization of the variation of sand probability in the study area would assist in identifying placement of potential wells for future drilling operations.

Software Aging Analysis of Web Server Using Neural Networks

Software aging is a phenomenon that refers to progressive performance degradation or transient failures or even crashes in long running software systems such as web servers. It mainly occurs due to the deterioration of operating system resource, fragmentation and numerical error accumulation. A primitive method to fight against software aging is software rejuvenation. Software rejuvenation is a proactive fault management technique aimed at cleaning up the system internal state to prevent the occurrence of more severe crash failures in the future. It involves occasionally stopping the running software, cleaning its internal state and restarting it. An optimized schedule for performing the software rejuvenation has to be derived in advance because a long running application could not be put down now and then as it may lead to waste of cost. This paper proposes a method to derive an accurate and optimized schedule for rejuvenation of a web server (Apache) by using Radial Basis Function (RBF) based Feed Forward Neural Network, a variant of Artificial Neural Networks (ANN). Aging indicators are obtained through experimental setup involving Apache web server and clients, which acts as input to the neural network model. This method is better than existing ones because usage of RBF leads to better accuracy and speed in convergence.