Ensemble learning techniques have achieved state-of-the-art performance in diverse machine learning applications by combining the predictions from two or more base models. This paper presents a ...concise overview of ensemble learning, covering the three main ensemble methods: bagging, boosting, and stacking, their early development to the recent state-of-the-art algorithms. The study focuses on the widely used ensemble algorithms, including random forest, adaptive boosting (AdaBoost), gradient boosting, extreme gradient boosting (XGBoost), light gradient boosting machine (LightGBM), and categorical boosting (CatBoost). An attempt is made to concisely cover their mathematical and algorithmic representations, which is lacking in the existing literature and would be beneficial to machine learning researchers and practitioners.
In recent years, the increased use of wireless networks for the transmission of large volumes of information has generated a myriad of security threats and privacy concerns; consequently, there has ...been the development of a number of preventive and protective measures including intrusion detection systems (IDS). Intrusion detection mechanisms play a pivotal role in securing computer and network systems; however, for various IDS, the performance remains a major issue. Moreover, the accuracy of existing methodologies for IDS using machine learning is heavily affected when the feature space grows. In this paper, we propose a IDS based on deep learning using feed forward deep neural networks (FFDNNs) coupled with a filter-based feature selection algorithm. The FFDNN-IDS is evaluated using the well-known NSL-knowledge discovery and data mining (NSL-KDD) dataset and it is compared to the following existing machine learning methods: support vectors machines, decision tree, K-Nearest Neighbor, and Naïve Bayes. The experimental results prove that the FFDNN-IDS achieves an increase in accuracy in comparison to other methods.
Many real-world machine learning applications require building models using highly imbalanced datasets. Usually, in medical datasets, the healthy patients or samples are dominant, making them the ...majority class, while the sick patients are few, making them the minority class. Researchers have proposed numerous machine learning methods to predict medical diagnosis. Still, the class imbalance problem makes it difficult for classifiers to adequately learn and distinguish between the minority and majority classes. Cost-sensitive learning and resampling techniques are used to deal with the class imbalance problem. This research focuses on developing robust cost-sensitive classifiers by modifying the objective functions of some well-known algorithms, such as logistic regression, decision tree, extreme gradient boosting, and random forest, which are then used to efficiently predict medical diagnosis. Meanwhile, as opposed to resampling techniques, our approach does not alter the original data distribution. Firstly, we implement the standard versions of these algorithms to provide a baseline for performance comparison. Secondly, we develop their corresponding cost-sensitive algorithms. For the proposed approaches, it is not necessary to change the distribution of the original data as the modified algorithms consider the imbalanced class distribution during training, thereby resulting in more reliable performance than when the data is resampled. Four popular medical datasets, including the Pima Indians Diabetes, Haberman Breast Cancer, Cervical Cancer Risk Factors, and Chronic Kidney Disease datasets, are used in the experiments to validate the performance of the proposed approach. The experimental results show that the cost-sensitive methods yield superior performance compared to the standard algorithms.
Full text
Available for:
GEOZS, IJS, IMTLJ, KILJ, KISLJ, NLZOH, NUK, OILJ, PNG, SAZU, SBCE, SBJE, UILJ, UL, UM, UPCLJ, UPUK, ZAGLJ, ZRSKP
The recent advances of e-commerce and e-payment systems have sparked an increase in financial fraud cases such as credit card fraud. It is therefore crucial to implement mechanisms that can detect ...the credit card fraud. Features of credit card frauds play important role when machine learning is used for credit card fraud detection, and they must be chosen properly. This paper proposes a machine learning (ML) based credit card fraud detection engine using the genetic algorithm (GA) for feature selection. After the optimized features are chosen, the proposed detection engine uses the following ML classifiers: Decision Tree (DT), Random Forest (RF), Logistic Regression (LR), Artificial Neural Network (ANN), and Naive Bayes (NB). To validate the performance, the proposed credit card fraud detection engine is evaluated using a dataset generated from European cardholders. The result demonstrated that our proposed approach outperforms existing systems.
Full text
Available for:
CEKLJ, NUK, ODKLJ, UL, UM, UPUK
Load frequency control or automatic generation control is one of the main operations that take place daily when considering a modern power system or not. The objectives of load frequency control are ...to maintain power balance between interconnected areas and to control the power flow in the tie-lines. Electric power cannot be stocked in large quantity that why its production must be equal to the consumption in each time. This equation constituted the key for a good management of any power system and introduces the need of more controllers when taking into account the integration of renewable energy sources into the traditional power system. More controllers are presented in the literature and the presents work proposes a review of those called traditional and those who combined the traditional controller and artificial intelligence algorithms.
Full text
Available for:
GEOZS, IJS, IMTLJ, KILJ, KISLJ, NLZOH, NUK, OILJ, PNG, SAZU, SBCE, SBJE, UILJ, UL, UM, UPCLJ, UPUK, ZAGLJ, ZRSKP
Credit cards play an essential role in today's digital economy, and their usage has recently grown tremendously, accompanied by a corresponding increase in credit card fraud. Machine learning (ML) ...algorithms have been utilized for credit card fraud detection. However, the dynamic shopping patterns of credit card holders and the class imbalance problem have made it difficult for ML classifiers to achieve optimal performance. In order to solve this problem, this paper proposes a robust deep-learning approach that consists of long short-term memory (LSTM) and gated recurrent unit (GRU) neural networks as base learners in a stacking ensemble framework, with a multilayer perceptron (MLP) as the meta-learner. Meanwhile, the hybrid synthetic minority oversampling technique and edited nearest neighbor (SMOTE-ENN) method is employed to balance the class distribution in the dataset. The experimental results showed that combining the proposed deep learning ensemble with the SMOTE-ENN method achieved a sensitivity and specificity of 1.000 and 0.997, respectively, which is superior to other widely used ML classifiers and methods in the literature.
Part of the widely discussed problem in electrical power systems is the optimal reactive power dispatch (ORPD) due to its reliability and economical operation of electrical power systems. The ORPD is ...a complex and nonlinear optimization problem. The pathfinder algorithm (PFA) is a newly developed algorithm that inspires the group movement of prey with a leader called a pathfinder when hunting for food. The inertia weight is added to the PFA and is called an improved pathfinder algorithm (IPFA) to support the proper random work of the swarm to avoid the decrease in searchability of the PFA. The IPFA was proposed in this work to diminish the active power loss while improving the voltage profile. The IPFA was validated on the IEEE 30 and 118 bus systems along with particle swarm optimization (PSO) and the teaching–learning-based optimizer (TLBO). The proposed IPFA provides the best result as the losses of the IEEE 30 and 118 test systems were reduced to 16.035 and 115.048 MW from the initial base of 17.89 and 132.86 MW, respectively. The losses of PSO and the TLBO were 16.1568 and 16.1607 MW for the IEEE 30 bus system, respectively, while for the IEEE 118 bus system, the PSO provided 117.9129 MW and the TLBO provided 118.0524 MW. The two test systems’ reduction percentages (%) were 10.37% and 13.41%, respectively. The results were compared with those of other algorithms in the literature, and the IPFA provided a superior result, thereby suggesting the superiority of IPFA methods in diminishing the power loss and improving the system’s voltage profile.
Full text
Available for:
IZUM, KILJ, NUK, PILJ, PNG, SAZU, UL, UM, UPUK
Algorithms are used to optimize both single and multi-objective system limits. This research aimed to detect the optimal location and size of the DGs, which can significantly minimize power loss and ...improve the stability of the voltage. The research uses binary particle swarm optimization and shuffled frog leap (BPSO-SLFA) algorithms for simulation and testing of an optimal power flow (OPF) on 33 and 69 bus radial distribution system. The result shows that the algorithms give better DG allocation and minimizes the power losses but at the nascent stage of advancement. The power losses associated with the system have significantly reduced up to 31.8244kW using multi-DGs reconfiguration placement. The outcomes are established to verify the potency of the recommend algorithm to minimize losses, general improvement in voltage profiles and cost saving for various distribution system. However, the proposed methodology can be used as a reliable method in DG settings and sizing in distribution network system which produce better outputs rather than hybrid grey wolf optimization (GWO) and hybrid big bang big crunch.
Full text
Available for:
GEOZS, IJS, IMTLJ, KILJ, KISLJ, NLZOH, NUK, OILJ, PNG, SAZU, SBCE, SBJE, UILJ, UL, UM, UPCLJ, UPUK, ZAGLJ, ZRSKP
Transactive energy trading, control, and management have gained much importance, receiving significant interest from both industry and academia. As a result, new strategies for its development and ...implementation are emerging. Prosumage, which combines household loads with solar PV and battery storage systems, has been presented as an effective technique for facilitating the integration of renewable energy sources while reducing distribution grid stress. Therefore, this paper proposes a grid-connected prosumer-centric residential smart community that uses cooperative game theory to exchange, regulate, and plan surplus energy generated by distributed energy resources (DERs) with the other neighbouring prosumers with energy deficits. The energy community considered in this work comprises ten identical prosumers (residential load, solar PV, and battery storage system), an energy community manager, and an energy retailer, all of whom are linked as a group to a distribution grid. The proposed market model seeks to examine the economic benefits of such renewable sources participating in centralized transactive energy trading. A centralized approach is adopted in the transactive energy trading among the ten proactive prosumers and the power grid through an energy community manager that administers trading activities inside the energy community. An optimization model is proposed in the centralized transactive energy trading market to optimize the financial benefits of rooftop solar PV-battery systems. The proposed model will not only optimally exchange, plan, and regulate the community domestic load, but it will also minimize total energy costs, increase system operation efficiency, and reduce system operating stress and carbon emissions.
Full text
Available for:
GEOZS, IJS, IMTLJ, KILJ, KISLJ, NLZOH, NUK, OILJ, PNG, SAZU, SBCE, SBJE, UILJ, UL, UM, UPCLJ, UPUK, ZAGLJ, ZRSKP
This paper investigates energy management systems in micro-grid using an optimization-based approach, optimizing the operating cost related to the energy purchased from the utility grid, the ...operation cost of the energy storage system, and revenue from the selling of energy to the utility grid. This research uses a constrained Particle Swarm Optimization-Based Model Predictive Control (CPSO-MPC) and a Linear Program-Based Optimization approach to solve the constrained optimization problem formulated in micro-grid energy management. Due to the absence of constraint management strategies in the traditional PSO algorithm, it is incapable of solving constrained optimization problems. Hence, to overcome this drawback, an intuitive approach known as Deb’s rule is applied to handle the constraints. The simulation results show the modified particle swarm optimization’s effective performance embedded in the model predictive control algorithm compared to the linear programming algorithm.
Full text
Available for:
GEOZS, IJS, IMTLJ, KILJ, KISLJ, NLZOH, NUK, OILJ, PNG, SAZU, SBCE, SBJE, UILJ, UL, UM, UPCLJ, UPUK, ZAGLJ, ZRSKP