Random Link Research Articles

Structured Sparse Regularization based Random Vector Functional Link Networks for DNA N4-methylcytosine sites prediction

As an epigenetic modification that plays an important role in modifying gene function and controlling gene expression during cell development, DNA N4-methylcytosine (4mC) is still lack of researching. It is therefore necessary to accurately predict the 4mC sites to make fully aware of its mechanism and function. In this paper, we propose a novel model which is called Structural Sparse Regularized Random Vector Functional Link Network (SSR-RVFL) for predicting 4mC sites. Compared with other state-of-the-art methods, SSR-RVFL performs better and achieves higher prediction accuracy. There are total six benchmark datasets used in the experiments, namely C.elegans, D.elanogaster, E.coli, A.thaliana G.subterraneus and G.pickeringii. Our model improves the accuracy by 0.42%, 0.45%, 0.48%, 0.91%, 0.66% and 0.7% on these six benchmark datasets respectively, so it can be regarded as a more effective prediction tool.

Research on improving robustness strategy of grid cyber physical system considering spatial information

Adding additional links can be recognized as an effective way to improve the robustness of grid cyber physical system (GCPS) against cascading failures. However, for a typical large-scale and across-regional GCPS, previous studies mainly focus on the network topology without considering the components/nodes spatial information. Therefore, to improve the robustness of GCPS without changing the existing connections, five different link addition strategies, considering spatial information and construction cost constraints, are proposed. Firstly, a spatial GCPS model is established, which consists of the spatial power system and the spatial information system with hierarchical structures. Among them, the spatial information system consists of the access layer (AL), the backbone layer (BL) and the core layer (CL) and its network nodes are divided into different layers according to their functions. Secondly, based on the spatial information system layer nodes functions and the spatial power system nodes attributions, five link addition strategies, AL-AL, AL-BL, AL-CL, Generators-CL (G-CL) and Loads-CL (L-CL), are investigated under two constraint scenarios. Thirdly, to verify our proposed five link addition strategies have better improvement effects, we compare them with the two existing excellent strategies, including low degree link addition strategy (LD) and random addition link addition strategy (Random). Finally, the statistical test results on the real 39-node power system and the IEEE 118 bus system show that our proposed link addition strategies not only effectively improve the robustness of GCPS, but also have better effects than LD and Random. More specifically, no matter whether GCPS suffered severe or slight attacks, the AL-BL always performs better when either fixed number constraint scenario or fixed length constraint scenario, and the improvement effects of AL-AL, AL-CL, G-CL and L-CL are exhibited as descending orders, respectively. These findings can help maintainers to improve the robustness of GCPS before attacks.

Estimation of compressive strength of concrete cement using random vector functional link networks: a case study

Estimation of compressive strength of concrete cement using random vector functional link networks: a case study

Modelling of thermo-hydraulic behavior of a helical heat exchanger using machine learning model and fire hawk optimizer

The study of the fluid flow in heat exchangers and its related mass and heat transfer processes is a very complicated research subject. Modeling of coupled fluid flow and heat transfer processes using numerical and analytical approaches are cumbersome problems. Hence, the current study investigates the modeling thermo-hydraulic behavior of a helical plate heat exchanger (HPHE) using advanced machine learning models. The outlet temperatures of hot and cold fluids are predicted considering different cross-sectional areas and pitches of the flow channels. This paper fulfills the research gap about fluid flow and heat transfer mechanisms by considering advanced machine learning approaches. The proposed model helps different manufacturers of heat exchangers to estimate the thermal performance and characteristics. The developed model aims to provide an advanced random vector functional link (RVFL) optimized fire hawk optimizer to predict outlet temperatures of working fluids of HPHE. The proposed model was compared with four optimized models using grey wolf, jellyfish, sine-cosine, and hybrid salp swarm-whale optimizers. The results of all models were compared and fire hawk optimizers showed superior accuracy compared with other models. Fire hawk optimizer had the highest R2 (0.999) followed by jellyfish (0.998) for both investigated responses. Hybrid salp swarm-whale and sine-cosine optimizers had the lowest R2 (0.987) in the case of hot fluid.

Online learning using deep random vector functional link network

Deep neural networks have shown their promise in recent years with their state-of-the-art results. Yet, backpropagation-based methods may suffer from time-consuming training process and catastrophic forgetting when performing online learning. In this work we attempt to curtail them by employing the ensemble deep Random Vector Functional Link (edRVFL). As opposed to backpropagation-based neural networks that adjust weights iteratively, RVFL uses a closed-form solution method without iterative parameter learning. In addition, our approach allows the model to grow incrementally as new data is made available so that it can more resemble real-life learning scenarios. Our proposed online learning models were able to perform better on 72% of the datasets in the classification scenario and 80% of the datasets in the regression scenario, when compared to other available randomization-based online learning models in the literature. This is further supported by statistical comparisons which also show the stability of our network.

DF-dRVFL: A novel deep feature based classifier for breast mass classification

Amongst all types of cancer, breast cancer has become one of the most common cancers in the UK threatening millions of people’s health. Early detection of breast cancer plays a key role in timely treatment for morbidity reduction. Compared to biopsy, which takes tissues from the lesion for further analysis, image-based methods are less time-consuming and pain-free though they are hampered by lower accuracy due to high false positivity rates. Nevertheless, mammography has become a standard screening method due to its high efficiency and low cost with promising performance. Breast mass, as the most palpable symptom of breast cancer, has received wide attention from the community. As a result, the past decades have witnessed the speeding development of computer-aided systems that are aimed at providing radiologists with useful tools for breast mass analysis based on mammograms. However, the main issues of these systems include low accuracy and require enough computational power on a large scale of datasets. To solve these issues, we developed a novel breast mass classification system called DF-dRVFL. On the public dataset DDSM with more than 3500 images, our best model based on deep random vector functional link network showed promising results through five-cross validation with an averaged AUC of 0.93 and an average accuracy of 81.71%\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$81.71\\%$$\\end{document}. Compared to sole deep learning based methods, average accuracy has increased by 0.38. Compared with the state-of-the-art methods, our method showed better performance considering the number of images for evaluation and the overall accuracy.

Association of white matter volume with brain age classification using deep learning network and region wise analysis

Structural magnetic resonance imaging (sMRI) has been used to examine age-related neuroanatomical changes in the human brain. In the present work, a pre-trained deep learning model and an ensemble deep random vector functional link (edRVFL) classifier have been used to create a brain age classification framework from magnetic resonance imaging (MRI) scans. A total of 155 MRI scans of the brain are obtained from the open-access OpenNeuro database and categorized into three age groups (3–5 years old, 7–12 years old, and 18–40 years old). To visualize the age connection across different brain regions, all MRI scans are first segmented into Gray Matter (GM), White Matter (WM), and Cerebrospinal Fluid (CSF). The ResNet-50 network is used to extract features from MRI images, while the edRVFL network is used to classify the retrieved features. Classification accuracy for GM, WM, CSF, and whole brain images are 96.11%, 98.33%, 93.33%, and 94.00%, respectively, using the edRVFL classifier. Region-wise analysis has also been done using Pearson’s correlation coefficient (r), coefficient of determination (R2), and root mean square error (RMSE) to analyze the relationship between brain age and brain tissue volumes. According to the findings of the suggested deep model for brain age categorization, and region-wise analysis, alterations in WM volume are strongly linked to brain aging.

Online dynamic ensemble deep random vector functional link neural network for forecasting

This paper proposes a three-stage online deep learning model for time series based on the ensemble deep random vector functional link (edRVFL). The edRVFL stacks multiple randomized layers to enhance the single-layer RVFL’s representation ability. Each hidden layer’s representation is utilized for training an output layer, and the ensemble of all output layers forms the edRVFL’s output. However, the original edRVFL is not designed for online learning, and the randomized nature of the features is harmful to extracting meaningful temporal features. In order to address the limitations and extend the edRVFL to an online learning mode, this paper proposes a dynamic edRVFL consisting of three online components, the online decomposition, the online training, and the online dynamic ensemble. First, an online decomposition is utilized as a feature engineering block for the edRVFL. Then, an online learning algorithm is designed to learn the edRVFL. Finally, an online dynamic ensemble method, which can measure the change in the distribution, is proposed for aggregating all layers’ outputs. This paper evaluates and compares the proposed model with state-of-the-art methods on sixteen time series.

Barycentric coordinate-based distributed localization for wireless sensor networks subject to random lossy links

Distributed localization in wireless sensor networks is essential for verifying the positions of the sensor nodes deployed in the sensing field based on information interactions. However, the information transmitted from the controller to the actuator for each sensor node may be lost owing to communication noise, channel interference, or network congestion. With this in mind, this paper aims at investigating the distributed localization for wireless sensor networks subject to random lossy links. The barycentric coordinate, which can be determined by the distance measurements between node pairs in the network, is used as the basic scheme for range-based localization. First, based on the characterization of random link loss in wireless sensor networks with a Bernoulli random variable, a distributed iterative localization algorithm is proposed. And then, the global convergence of the proposed localization algorithm which ensures accurate localization is theoretically proved by using the convergence of sub-stochastic matrices’ product. Finally, numerical examples are conducted to illustrate the effectiveness of the proposed localization algorithm.

CovC-ReDRNet: A Deep Learning Model for COVID-19 Classification.

Since the COVID-19 pandemic outbreak, over 760 million confirmed cases and over 6.8 million deaths have been reported globally, according to the World Health Organization. While the SARS-CoV-2 virus carried by COVID-19 patients can be identified though the reverse transcription-polymerase chain reaction (RT-PCR) test with high accuracy, clinical misdiagnosis between COVID-19 and pneumonia patients remains a challenge. Therefore, we developed a novel CovC-ReDRNet model to distinguish COVID-19 patients from pneumonia patients as well as normal cases. ResNet-18 was introduced as the backbone model and tailored for the feature representation afterward. In our feature-based randomized neural network (RNN) framework, the feature representation automatically pairs with the deep random vector function link network (dRVFL) as the optimal classifier, producing a CovC-ReDRNet model for the classification task. Results based on five-fold cross-validation reveal that our method achieved 94.94%, 97.01%, 97.56%, 96.81%, and 95.84% MA sensitivity, MA specificity, MA accuracy, MA precision, and MA F1-score, respectively. Ablation studies evidence the superiority of ResNet-18 over different backbone networks, RNNs over traditional classifiers, and deep RNNs over shallow RNNs. Moreover, our proposed model achieved a better MA accuracy than the state-of-the-art (SOTA) methods, the highest score of which was 95.57%. To conclude, our CovC-ReDRNet model could be perceived as an advanced computer-aided diagnostic model with high speed and high accuracy for classifying and predicting COVID-19 diseases.

Contribution of Fluid Substitution and Cheetah Optimizer Algorithm in Predicting Rock-Physics Parameters of Gas-Bearing Reservoirs in the Eastern Mediterranean Sea, Egypt

AbstractIn this study, the elastic characteristics of reservoir rocks and their relationship to porosity and pore fluid were predicted using the fluid substitution method in combination with machine learning techniques. We first discarded the data at gas points to remove the erroneous effect of gas on the prediction process of Poisson’s ratio using the three proposed machine learning models. Then, the prediction was carried out after substituting the gas zones by oil and by water. As a result, the prediction was enhanced and showed stronger correlation coefficient values. The integration of fluid substitution and machine learning methods was applied in the reservoir of Scarab field as a case study from the Eastern Mediterranean to detect the effect of different pore fluids (gas, oil, and water) on Poisson's ratio estimation. The main objective of the study was to analyze the seismic and well log data to estimate and predict the Poisson’s ratio in four fluid-content cases; these are gas-bearing reservoir, reservoir after removal of log data of gas-bearing zones, and reservoirs after gas-substitution with oil and with water. These four cases were dealt with directly and by using the machine learning algorithms based on the proposed model of random vector functional link (RVFL), which was enhanced by the Cheetah optimizer (CO). This study shows how the performance of RVFL is affected by the presence or absence of gas zones. It is shown that the Poisson’s ratio value increases when gas is substituted with water more than when gas is substituted with oil. For validation of these results, regression analysis technique was used and the correlation coefficient of the CO–RVFL model increased after removing well log data of gas zones and was more enhanced after fluid substitution from gas to oil or to water.

An exact approach for finding bicriteria maximally SRLG-disjoint/shortest path pairs in telecommunication networks

The paper addresses a bicriteria optimisation problem in telecommunication networks that aims at finding Pareto efficient pairs of paths between two given nodes, seeking to minimise the number of SRLGs (Shared Risk Link Groups) common to both paths and the path pair cost. This problem is of particular importance in telecommunication routing design, namely concerning resilient routing models where both a primary and a backup paths have to be calculated to minimise the risk of failure of a connection between origin and terminal nodes, in case of failure in the primary path. An exact resolution method is applied for solving this problem, enabling the calculation of the whole set of Pareto optimal solutions, which combines a transformation of the network representation with a path ranking algorithm. A comprehensive experimental study on the application of this approach, using reference network topologies, considering random SRLG assignments to the links and random link bandwidth occupations, together with the discussion on typical examples of solution selection and potential advantages of the method, are presented.

Gray Level Co-Occurrence Matrix and RVFL for Covid-19 Diagnosis

As the widespread transmission of COVID-19 has continued to influence human health since late 2019, more intersections between artificial intelligence and the medical field have arisen. For CT images, manual differentiation between COVID-19-infected and healthy control images is not as effective and fast as AI. This study performed experiments on a dataset containing 640 samples, 320 of which were COVID-19-infected, and the rest were healthy controls. This experiment combines the gray-level co-occurrence matrix (GLCM) and random vector function link (RVFL). The role of GLCM and RVFL is to extract image features and classify images, respectively. The experimental results of my proposed GLCM-RVFL model are validated using K-fold cross-validation, and the indicators are 78.81±1.75%, 77.08±0.68%, 77.46±0.73%, 54.22±1.35%, and 77.48±0.74% for sensitivity, accuracy, F1-score, MCC, and FMI, respectively, which also confirms that the proposed model performs well on the COVID-19 detection task. After comparing with six state-of-the-art COVID-19 detection, I ensured that my model achieved higher performance.

A stochastic configuration networks based on Harris hawks optimizer

Stochastic configuration networks (SCNs), as a novel incremental generation model with supervisory mechanism, has an excellent superiority in solving large-scale data regression and classification problems. However, the accuracy of the SCNs is affected by the assignation and selection of some network parameters significantly. Harris hawk optimizer (HHO) algorithm is a new meta-heuristic algorithm that simulates the foraging and anti-predation behavior of Harris hawk. In this paper, a SCNs based on HHO algorithm is first introduced, termed as HHO-SCNs. As the performance of SCNs is related to regularization parameter r and scale factor lambda of weights and biases, then HHO is employed to give better parameters for SCNs automatically. A numerical function and six benchmark datasets are used to verify the regression performance of the proposed model. Three benchmark datasets are introduced to illustrate the effectiveness of the proposed model for classification performance. Experimental results demonstrate the feasibility and validity of HHO-SCNs compared with incremental random vector functional link, SCNs, fast SCNs, and SCNs based on whale optimization algorithm. The proposed HHO-SCNs improves the generalization performance of standard SCNs, and provides a new idea for expanding the development and application of SCNs.

Random vector functional link network: Recent developments, applications, and future directions

Neural networks have been successfully employed in various domains such as classification, regression and clustering, etc. Generally, the back propagation (BP) based iterative approaches are used to train the neural networks, however, it results in the issues of local minima, sensitivity to learning rate and slow convergence. To overcome these issues, randomization based neural networks such as random vector functional link (RVFL) network have been proposed. RVFL model has several characteristics such as fast training speed, direct links, simple architecture, and universal approximation capability, that make it a viable randomized neural network. This article presents the first comprehensive review of the evolution of RVFL model, which can serve as the extensive summary for the beginners as well as practitioners. We discuss the shallow RVFLs, ensemble RVFLs, deep RVFLs and ensemble deep RVFL models. The variations, improvements and applications of RVFL models are discussed in detail. Moreover, we discuss the different hyperparameter optimization techniques followed in the literature to improve the generalization performance of the RVFL model. Finally, we present potential future research directions/opportunities that can inspire the researchers to improve the RVFL’s architecture and learning algorithm further.

Two-step approach based multi-objective groundwater remediation using enhanced random vector functional link integrated with evolutionary marine predator algorithm

We here propose a two-step approach-based simulation-optimization model for multi-objective groundwater remediation using enhanced random vector functional link (ERVFL) and evolutionary marine predator algorithm (EMPA). In this study, groundwater flow and solute transport models are developed using MODFLOW and MT3DMS. The ERVFL network is used to approximate the flow and transport models, enhancing the computational performance. This study also improves the robustness of the ERVFL network using a kernel density estimator (KDE) based weighted least square approach. We further develop the EMPA by modifying the marine predator algorithm (MPA) using elite opposition-based learning, biological evolution operators, and elimination mechanisms. In the multi-objective version of EMPA, the non-dominated/Pareto-optimal solutions are stored in an external repository using an archive controller and adaptive grid mechanism to promote better convergence and diversity of the Pareto front. The proposed methodologies are applied for multi-objective groundwater remediation of a hypothetical unconfined aquifer based on the two-step method. The first step directly integrates flow and transport models with EMPA and finds the optimal locations of pumping wells by minimizing the percent of contaminant mass remaining in the aquifer. In the second step, the ERVL-based proxy model is integrated with EMPA and used for multi-objective optimization while explicitly using the pumping well locations obtained in the first step. The multi-objective optimization generates a Pareto-optimal solution representing the relationship between the rate of pumping and the amount of contaminant mass in the aquifer. Further analyses show a significant advantage of the two-step approach over a traditional method for multi-objective groundwater remediation.

Wireless Network Slice Assignment With Incremental Random Vector Functional Link Network

This paper presents an artificial intelligence-assisted network slice prediction method, which utilizes a novel incremental random vector functional link (IRVFL) network to deal with the wireless network slice assignment (WNSA) problem in a data-driven way. The goal of WNSA is to assign an appropriate network slice for a user's requirement based on the next-generation wireless derive and communication data. The IRVFL network is an incremental version of the RVFL network, where a data stream processing approach is used to gradually update output layer weights as new data arrive rather than processing the data as a single large data set. To ensure that the RVFL network can be trained for WNSA and have high adaptability and expansibility, we derive a novel flexible and appropriate rule for updating output layer weights of the IRVFL network. We have carried out extensive experiments to validate the feasibility, rationality, and effectiveness of using the IRVFL network for the WNSA problem. Results show that network slice prediction converges as the IRVFL network is incrementally trained and that the time required for training the incremental RVFL network is far less than for the non-incremental RVFL network. The incremental training of IRVFL network improves the performance of wireless network slice prediction. In addition, a comparison with six classification algorithms reveal that the IRVFL network consumes the least amount of time and has equivalent wireless network slice prediction performance.

Ensemble Deep Random Vector Functional Link Neural Network for Regression

Inspired by the ensemble strategy of machine learning, deep random vector functional link (dRVFL), and ensemble dRVFL (edRVFL) has shown state-of-the-art results on different datasets. Our present work first fills the gap of dRVFL and edRVFL work in the field of regression. We test and evaluate the performances of the dRVFLs on regression problems. Subsequently, we propose a novel regularization method boosted factor (BF), two dRVFLs variants edRVFL with skip connection (edRVFL-SC) and edRVFL with random skip connections (edRVFL-RSC) and one strategy ensemble skip connection edRVFL (esc-edRVFL) which show significant improvement over the original dRVFL. The BF is a newly introduced hyperparameter to scale the values of the activated hidden neurons to accommodate the diversity of the data, and it is also able to filter the neurons. edRVFL-SC and edRVFL-RSC are the edRVFL variants with skip connections. In edRVFL-SC, we apply dense skip connections to the edRVFL, which is inspired by the residual architecture in the deep learning area. However, due to the specificity of randomized networks, the simple skip connections are probably leading to the reuse of useless features. To address this problem, we propose a random skip connection-based edRVFL, which can keep the diversity in the latent space. esc-RVFL is an ensemble scheme that utilizes several edRVFL-RSC models trained on the different folds of the training dataset. The esc-edRVFL is identified as the best-performing algorithm through a comprehensive evaluation of 31 UCI datasets.

Prediction of the tensile properties of ultrafine grained Al–SiC nanocomposites using machine learning

We discovered and analyzed the new prediction model by using machine learning (ML) for the tensile strength of aluminum nanocomposites reinforced with μ-SiC particles fabricated by accumulative roll bonding (ARB). The effect of the number of cycles and SiC content on the microstructure, phase analysis, tensile, and hardness properties have been investigated for the ARBed sheets and their composites. The experimental results showed the distribution of SiC particles improved by increasing ARB passes. The ARB approach greatly enhanced the ultimate tensile strength (UTS), yield strength (YS), and hardness. The UTS achieved was 254 MPa for 4% SiC after 9 ARB cycles. The hardness values of the ARBed AA1050, and AA1050-4 wt% SiC are 60, and 76.5, respectively, after 9 ARB cycles. The modified version of random vector functional link based on Growth Optimizer Algorithm is developed as a machine-learning model to predict the tensile properties of the produced composites. The efficiency of the developed ML model is evaluated with other methods according to the performance criteria.

Chimeras in complex networks: A gear by nonlinear mean-field

We investigate a network of oscillators with a nonlinear memristive mean-field along with non-local interaction. We found the culmination of interesting chimeric patterns. With the change in the range of interaction the symmetric breaking bifurcation lead to chimeras. We derive analytically the critical condition for the appearance of chimera in the mean-field limit. The chimeras are characterized by the average of the local curvature of the node. We also explored the emergence of chimeric patterns in a network by replacing the non-local topology with random links. We found that the clustering of the nearby state to the solitary state causes the chimeras in random configuration with a memristive mean-field. The findings are illustrated through numerical simulations.