Search results

The purpose of this paper is to propose successive-over-relaxation (SOR) based recursive Bayesian approach (RBA) for the configuration identification of a Power System. Moreover, to present a comparison between the proposed method and existing RBA approaches regarding convergence speed and robustness.

Swift power network configuration identification is important for adopting the smart grid features like power system automation. In this work, a new SOR-based numerical approach is adopted to increase the convergence speed of the existing RBA algorithm and at the same time maintaining robustness against noise. Existing RBA and SOR-RBA are tested on IEEE 6 bus, IEEE 14 bus networks and 48 bus Danish Medium Voltage distribution network in the MATLAB R2014b environment and a comparative analysis is presented.

The comparison of existing RBA and proposed SOR-RBA is performed, which reveals that the latter has good convergence speed compared to the former RBA algorithms. Moreover, it is robust against bad data and noise.

Existing RBA techniques have slow convergence and are also prone to measurement noise. Their convergence speed is effected by noisy measurements. In this paper, an attempt has been made to enhance convergence speed of the new identification algorithm while keeping its numerical stability and robustness during noisy measurement conditions. This work is novel and has drastic improvement in the convergence speed and robustness of the former RBA algorithms.

The purpose of this paper is to demonstrate the applicability of swarm and evolutionary techniques for regularized machine learning. Generally, by defining a proper penalty function, regularization laws are embedded into the structure of common least square solutions to increase the numerical stability, sparsity, accuracy and robustness of regression weights. Several regularization techniques have been proposed so far which have their own advantages and disadvantages. Several efforts have been made to find fast and accurate deterministic solvers to handle those regularization techniques. However, the proposed numerical and deterministic approaches need certain knowledge of mathematical programming, and also do not guarantee the global optimality of the obtained solution. In this research, the authors propose the use of constraint swarm and evolutionary techniques to cope with demanding requirements of regularized extreme learning machine (ELM).

To implement the required tools for comparative numerical study, three steps are taken. The considered algorithms contain both classical and swarm and evolutionary approaches. For the classical regularization techniques, Lasso regularization, Tikhonov regularization, cascade Lasso-Tikhonov regularization, and elastic net are considered. For swarm and evolutionary-based regularization, an efficient constraint handling technique known as self-adaptive penalty function constraint handling is considered, and its algorithmic structure is modified so that it can efficiently perform the regularized learning. Several well-known metaheuristics are considered to check the generalization capability of the proposed scheme. To test the efficacy of the proposed constraint evolutionary-based regularization technique, a wide range of regression problems are used. Besides, the proposed framework is applied to a real-life identification problem, i.e. identifying the dominant factors affecting the hydrocarbon emissions of an automotive engine, for further assurance on the performance of the proposed scheme.

Through extensive numerical study, it is observed that the proposed scheme can be easily used for regularized machine learning. It is indicated that by defining a proper objective function and considering an appropriate penalty function, near global optimum values of regressors can be easily obtained. The results attest the high potentials of swarm and evolutionary techniques for fast, accurate and robust regularized machine learning.

The originality of the research paper lies behind the use of a novel constraint metaheuristic computing scheme which can be used for effective regularized optimally pruned extreme learning machine (OP-ELM). The self-adaption of the proposed method alleviates the user from the knowledge of the underlying system, and also increases the degree of the automation of OP-ELM. Besides, by using different types of metaheuristics, it is demonstrated that the proposed methodology is a general flexible scheme, and can be combined with different types of swarm and evolutionary-based optimization techniques to form a regularized machine learning approach.

The purpose of this paper is to conduct a comprehensive review of the noteworthy contributions made in the area of the Feedforward neural network (FNN) to improve its generalization performance and convergence rate (learning speed); to identify new research directions that will help researchers to design new, simple and efficient algorithms and users to implement optimal designed FNNs for solving complex problems; and to explore the wide applications of the reviewed FNN algorithms in solving real-world management, engineering and health sciences problems and demonstrate the advantages of these algorithms in enhancing decision making for practical operations.

The FNN has gained much popularity during the last three decades. Therefore, the authors have focused on algorithms proposed during the last three decades. The selected databases were searched with popular keywords: “generalization performance,” “learning rate,” “overfitting” and “fixed and cascade architecture.” Combinations of the keywords were also used to get more relevant results. Duplicated articles in the databases, non-English language, and matched keywords but out of scope, were discarded.

The authors studied a total of 80 articles and classified them into six categories according to the nature of the algorithms proposed in these articles which aimed at improving the generalization performance and convergence rate of FNNs. To review and discuss all the six categories would result in the paper being too long. Therefore, the authors further divided the six categories into two parts (i.e. Part I and Part II). The current paper, Part I, investigates two categories that focus on learning algorithms (i.e. gradient learning algorithms for network training and gradient-free learning algorithms). Furthermore, the remaining four categories which mainly explore optimization techniques are reviewed in Part II (i.e. optimization algorithms for learning rate, bias and variance (underfitting and overfitting) minimization algorithms, constructive topology neural networks and metaheuristic search algorithms). For the sake of simplicity, the paper entitled “Machine learning facilitated business intelligence (Part II): Neural networks optimization techniques and applications” is referred to as Part II. This results in a division of 80 articles into 38 and 42 for Part I and Part II, respectively. After discussing the FNN algorithms with their technical merits and limitations, along with real-world management, engineering and health sciences applications for each individual category, the authors suggest seven (three in Part I and other four in Part II) new future directions which can contribute to strengthening the literature.

The FNN contributions are numerous and cannot be covered in a single study. The authors remain focused on learning algorithms and optimization techniques, along with their application to real-world problems, proposing to improve the generalization performance and convergence rate of FNNs with the characteristics of computing optimal hyperparameters, connection weights, hidden units, selecting an appropriate network architecture rather than trial and error approaches and avoiding overfitting.

This study will help researchers and practitioners to deeply understand the existing algorithms merits of FNNs with limitations, research gaps, application areas and changes in research studies in the last three decades. Moreover, the user, after having in-depth knowledge by understanding the applications of algorithms in the real world, may apply appropriate FNN algorithms to get optimal results in the shortest possible time, with less effort, for their specific application area problems.

The existing literature surveys are limited in scope due to comparative study of the algorithms, studying algorithms application areas and focusing on specific techniques. This implies that the existing surveys are focused on studying some specific algorithms or their applications (e.g. pruning algorithms, constructive algorithms, etc.). In this work, the authors propose a comprehensive review of different categories, along with their real-world applications, that may affect FNN generalization performance and convergence rate. This makes the classification scheme novel and significant.

The purpose of this paper is to introduce a new approach for the identification of price thresholds, which enables learning true thresholds from previous buying decisions recorded in POS scanner data.

The methodology presented herein combines spline regression with generalized cross‐validation. Classical Chi‐square testing confirms the separation of regimes of the price response function by this methodology. Five propositions concerning the consumers' response to odd pricing in a Western‐type market are evaluated.

Despite the widespread retail practice odd prices are unlikely to flag the actual threshold in consumer response. The term odd price refers to prices with a non‐zero ending in the cent digit, e.g. .95, .98 or .99, which are commonly used in Western‐type markets. Moreover, the simple odd price effects are distinguished from odd‐ending prices with the first number left of the decimal point set to an odd number. The results show that even these prices not always flag a threshold in consumer response.

The discussion of the odd‐price effect is confused by conflicting empirical results and related interpretations of the underlying mechanisms. In contrast to many previous investigations – which are restricted to the consideration of very few price endings – this study covers all reasonable prices. Statistically significant odd‐price effects are found for some brands, but not for all within the same category.

One must argue for checking the thresholds for each brand individually rather than generalizing by applying misleading rules of thumb.

The paper provides researchers as well as practitioners with a methodology to evaluate price thresholds and outlines the shortcoming of contemporary retailers pricing practices in a detailed manner.

Electrical capacitance tomography (ECT) and electrical resistance tomography (ERT) are promising techniques for multiphase flow measurement due to their high speed, low cost, non-invasive and visualization features. There are two major difficulties in image reconstruction for ECT and ERT: the “soft-field”effect, and the ill-posedness of the inverse problem, which includes two problems: under-determined problem and the solution is not stable, i.e. is very sensitive to measurement errors and noise. This paper aims to summarize and evaluate various reconstruction algorithms which have been studied and developed in the word for many years and to provide reference for further research and application.

In the past 10 years, various image reconstruction algorithms have been developed to deal with these problems, including in the field of industrial multi-phase flow measurement and biological medical diagnosis.

This paper reviews existing image reconstruction algorithms and the new algorithms proposed by the authors for electrical capacitance tomography and electrical resistance tomography in multi-phase flow measurement and biological medical diagnosis.

The authors systematically summarize and evaluate various reconstruction algorithms which have been studied and developed in the word for many years and to provide valuable reference for practical applications.

This paper aims to investigate the impact of oil price shocks on the Turkish sovereign yield curve factors.

To extract the latent factors (level, slope and curvature) of the Turkish sovereign yield curve, we estimate conventional Nelson and Siegel (1987) model with nonlinear least squares. Then, we decompose oil price shocks into supply, demand and risk shocks using structural VAR (structural VAR) models. After this separation, we apply Engle (2002) dynamic conditional correlation GARCH (DCC-GARCH (1,1)) method to investigate time-varying co-movements between yield curve factors and oil price shocks. Finally, using the LP (local projections) proposed by Jorda (2005), we estimate the impulse-response functions to examine the impact of different oil price shocks on yield curve factors.

Our results demonstrate that the various oil price shocks influence the yield curve factors quite differently. A supply shock leads to a statistically significant increase in the level factor. This result shows that elevated oil prices due to supply disruptions are interpreted as a signal of a surge in inflation expectations since the cost channel prevails. Besides, unanticipated demand shocks have a positive impact on the slope factor as a result of the central bank policy response for offsetting the elevated inflation expectations. Finally, a risk shock is associated with a decrease in the curvature factor indicating that risk shocks influence the medium-term bonds due to the deflationary pressure resulting from depressed economic conditions.

Our results provide new insights to understand the driving forces of yield curve movements induced by various oil shocks to formulate appropriate policy responses.

The study contributes to the literature by two main dimensions. First, the recent oil shock identification scheme of Ready (2018) is modified using the “geopolitical oil price risk index” to capture the changes in the risk perceptions of oil markets driven by geopolitical tensions such as terrorism and conflicts and sanctions. The modified identification scheme attributes more power to demand shocks in explaining the variation of the oil price compared to that of the baseline scheme. Second, it provides recent evidence that distinguishes the impact of oil demand and supply shocks on Turkey's yield curve.

Trust is one of the main pillars of many communication and interaction domains. Computing is no exception. Fog computing (FC) has emerged as mitigation of several cloud computing limitations. However, selecting a trustworthy node from the fog network still presents serious challenges. This paper aims to propose an algorithm intended to mitigate the trust and the security issues related to selecting a node of a fog network.

The proposed model/algorithm is based on two main concepts, namely, machine learning using fuzzy neural networks (FNNs) and the weighted weakest link (WWL) algorithm. The crux of the proposed model is to be trained, validated and used to classify the fog nodes according to their trust scores. A total of 2,482 certified computing products, in addition to a set of nodes composed of multiple items, are used to train, validate and test the proposed model. A scenario including nodes composed of multiple computing items is designed for applying and evaluating the performance of the proposed model/algorithm.

The results show a well-performing trust model with an accuracy of 0.9996. Thus, the end-users of FC services adopting the proposed approach could be more confident when selecting elected fog nodes. The trained, validated and tested model was able to classify the nodes according to their trust level. The proposed model is a novel approach to fog nodes selection in a fog network.

Certainly, all data could be collected, however, some features are very difficult to have their scores. Available techniques such as regression analysis and the use of the experts have their own limitations. Experts might be subjective, even though the author used the fuzzy group decision-making model to mitigate the subjectivity effect. A methodical evaluation by specialized bodies such as the security certification process is paramount to mitigate these issues. The author recommends the repetition of the same study when data form such bodies is available.

The novel combination of FNN and WWL in a trust model mitigates uncertainty, subjectivity and enables the trust classification of complex FC nodes. Furthermore, the combination also allowed the classification of fog nodes composed of diverse computing items, which is not possible without the WWL. The proposed algorithm will provide the required intelligence for end-users (devices) to make sound decisions when requesting fog services.

The main purpose of this paper is to predicate financial distress using the worst-practice-frontier data envelopment analysis (WPF-DEA) model and artificial neural network.

In this study, a neural network technique was used to forecast inputs and outputs in the future time-period. Using a WPF-DEA model, financially distressed companies were identified based on the worst performance, and an improvement solution was provided for those decision-making units.

This study’s findings show that dynamic WPF-DEA has high predictability in corporate financial distress, and it can be used with high confidence. Based on the future time-period results, JOUSH & OXYGEN was predicted to be a financially distressed company in the two future time-periods.

In recent decades, globalization, technological changes and a competitive space have increased uncertainty in the economic environment. In such circumstances, economic growth certainly depends on correct decision-making and optimal allocation of resources. It can be done by introducing appropriate tools and models for assessing corporate financial conditions, including financial distress and bankruptcy.

This paper aims to examine Nigeria’s dynamic output and output volatility connectedness with USA, China and India using quarterly data from 1981Q1 to 2019Q4.

The study adopted the network approach of Diebold and Yilmaz (2014) and used the normalized generalized forecast error variance decomposition from an underlying vector error correction model to build connectedness measures.

The findings show that the global financial crisis (GFC) increased the connectedness index far more than the 2016 Nigeria economic recession. The moderate effect of the 2016 Nigeria economic recession on the connectedness index underscores the fact that Nigeria is a small, open economy with minimal capacity to spread output shock. For both real output and its volatility, the total connectedness index rose smoothly and systematically through time, thereby leaving the economies more connected in the long run.

To the best of the authors’ knowledge, this paper is among the first to examine Nigeria’s dynamic output and output volatility connectedness with the USA, China and India using new empirical insights from the GFC versus 2016 Nigerian recession. The study, therefore, concludes that the Nigerian economy should be diversified immediately as a hedge against future real output shocks, while the USA, China and India should maintain and sustain their current policy frameworks to remain less vulnerable to real output shocks.

This paper aims to deal with the problem of the global stabilization for a class of tension leg platform (TLP) nonlinear control systems.

It is well-known that, in general, the global asymptotic stability of the TLP subsystems does not imply the global asymptotic stability of the composite closed-loop system.

An effective approach is proposed to control chaos via the combination of fuzzy controllers, fuzzy observers and dithers.

If a fuzzy controller and a fuzzy observer cannot stabilize the chaotic system, a dither, as an auxiliary of the controller and the observer, is simultaneously introduced to asymptotically stabilize the chaotic system.

Thus, the behavior of the closed-loop dithered chaotic system can be rigorously predicted by establishing that of the closed-loop fuzzy relaxed system.