Search

The integration of metaheuristics with machine learning methodologies presents significant advantages, particularly in optimization and computational intelligence. This amalgamation leverages the global search capabilities of metaheuristics alongside the pattern recognition and predictive prowess of machine learning, facilitating enhanced convergence rates and solution quality in complex problem spaces. The Quantum Long Short-Term Memory (QLSTM) emerges as a highly efficient deep learning model tailored to tackle such intricate engineering problems. The QLSTM's architecture, comprising data encoding, variational, and quantum measurement layers, facilitates the effective encoding and processing of civil engineering data, leading to heightened prediction accuracy. However, the task of determining optimal values for QLSTM parameters presents challenges due to its NP-problem nature and time-consuming characteristics. To address this, we propose an alternative technique to optimize the QLSTM based on a modified Electric Eel Foraging Optimization (MEEFO). The MEEFO is a modified version of the original EEFO that applies triangular mutation operators to boost the search capability of the traditional EEFO. Thus, the MEEFO optimizes the QLSTM and boosts its prediction performance. To validate the efficacy of our proposed method, we conduct comprehensive experiments utilizing five real-world engineering datasets related to construction and structure engineering. The evaluation outcomes unequivocally demonstrate that the MMEFO significantly enhances the performance of the QLSTM. [ABSTRACT FROM AUTHOR]

• Developing a new paradigm for the feature selection problem based on the hyperheuristic methodology. • Configuring the combination of the components in to automatically find the most pertinent features for each dataset. • Evaluating the performance of the proposed technique using a comprehensive set of eighteen datasets. • Comparing the results of the best FS combination with the other well-known FS methods. Feature selection (FS) is a crucial step for effective data mining since it has largest effect on improving the performance of classifiers. This is achieved by removing the irrelevant features and using only the relevant features. Many metaheuristic approaches exist in the literature in attempt to address this problem. The performance of these approaches differ based on the settings of a number of factors including the use of chaotic maps, opposition-based learning (OBL) and the percentage of the population that OBL will be applied to, the metaheuristic (MH) algorithm adopted, the classifier utilized, and the threshold value used to convert real solutions to binary ones. However, it is not an easy task to identify the best settings for these different components in order to determine the relevant features for a specific dataset. Moreover, running extensive experiments to fine tune these settings for each and every dataset will consume considerable time. In order to mitigate this important issue, a hyper-heuristic based FS paradigm is proposed. In the proposed model, a two-stage approach is adopted to identify the best combination of these components. In the first stage, referred to as the training stage , the Differential Evolution (DE) algorithm is used as a controller for selecting the best combination of components to be used by the second stage. In the second stage, referred to as the testing stage , the received combination will be evaluated using a testing set. Empirical evaluation of the proposed framework is based on numerous experiments performed on the most popular 18 datasets from the UCI machine learning repository. Experimental results illustrates that the generated generic configuration provides a better performance than eight other metaheuristic algorithms over all performance measures when applied to the UCI dataset. Moreover, The overall paradigm ranks at number one when compared against state-of-the-art algorithms. Finally, the generic configuration provides a very competitive performance for high dimensional datasets. [ABSTRACT FROM AUTHOR]

Quantitative Structure Activity Relation (QSAR) models are mathematical techniques used to link structural characteristics with biological activities, thus considered a useful tool in drug discovery, hazard evaluation, and identifying potentially lethal molecules. The QSAR regulations are determined by the Organization for Economic Cooperation and Development (OECD). QSAR models are helpful in discovering new drugs and chemicals to treat severe diseases. In order to improve the QSAR model's predictive power for biological activities of naturally occurring indoloquinoline derivatives against different cancer cell lines, a modified machine learning (ML) technique is presented in this paper. The Arithmetic Optimization Algorithm (AOA) operators are used in the suggested model to enhance the performance of the Sinh Cosh Optimizer (SCHO). Moreover, this improvement functions as a feature selection method that eliminates superfluous descriptors. An actual dataset gathered from previously published research is utilized to evaluate the performance of the suggested model. Moreover, a comparison is made between the outcomes of the suggested model and other established methodologies. In terms of pIC50 values for different indoloquinoline derivatives against human MV4-11 (leukemia), human HCT116 (colon cancer), and human A549 (lung cancer) cell lines, the suggested model achieves root mean square error (RMSE) of 0.6822, 0.6787, 0.4411, and 0.4477, respectively. The biological application of indoloquinoline derivatives as possible anticancer medicines is predicted with a high degree of accuracy by the suggested model, as evidenced by these findings. [Display omitted] • The primary contributions of this work can be summed up as follows: • Modified Sinh Cosh Optimizer (SCO)-QSAR model to predict pIC 50 values for various cells. • Improve the performance of SCO using Arithmetic Optimization Algorithm operators. • Assess the efficiency of the developed model using real-world QSAR datasets. [ABSTRACT FROM AUTHOR]

In this paper, an improved version of the Grey Wolf Optimizer (GWO) is proposed to improve the exploration and the exploitation ability of the GWO algorithm. This improvement is performed through using the chaotic logistic map, the Opposition-Based Learning (OBL), the differential evolution(DE), and the disruption operator (DO). Where, the chaotic logistic map and the OBL are used to initialize the candidate solutions and these approaches avoid the drawbacks of the random population and increase the convergence of the algorithm. Then, the DE operators are combined with the GWO algorithm, in which, the DE operators work as a local search mechanism to improve the exploitation ability of the GWO through updating the population. Also, after updating the solutions by using a hybrid between the GWO and the DE, the DO is used to enhance the exploration ability, in which, the DO is used to maintain the diversity of the population. Therefore, the combinations with chaotic logistic map, OBL, DE, and DO, provide the GWO with tools to better balance between the exploration and the exploitation of the search space without affecting the computational time required for this task. The proposed algorithm, called COGWO2D, is compared with other seven algorithms through a set of experimental series that have been performed over two benchmark functions, the classical CEC2005, and the CEC2014. Also, the performance of the proposed algorithm to improve the classification of the galaxy images is evaluated, where it is used as a feature selection method. The aim of this experiment is to select the optimal subset of features from the extracted features of the galaxy images. The experimental results support the efficacy of the proposed approach to find the optimal solutions of the global optimization problem, as well as, increase the accuracy of the classification of the galaxy images. [ABSTRACT FROM AUTHOR]

[Display omitted] • Developed 12 novel methanol/diesel blends, achieving up to 9.3% increase in BP. • Lowest BSFC of 0.27 kg/kWh in methanol/diesel blends, outperforming pure diesel. • Machine learning model has a prediction accuracy of R2 ≈ 93% and RMSE ≈ 1.13. • BTE increased by 31.5% with C2 blend, showing enhanced combustion efficiency. • Methanol/diesel blends showed stable VE between 71.96% and 76.65% across loads. The pursuit of reducing diesel consumption while progressing towards a sustainable energy future necessitates critical decisions regarding fuel modifications or engine adaptations to ensure smooth transitions in transportation. This study explores the potential of methanol/diesel blends as a sustainable fuel solution for the transport sector. We address a significant gap by examining the impact of six different surfactants on blend stability and engine performance. Ternary phase diagrams were constructed to analyse blend stability, and engine testing on a 3.5 kW single-cylinder diesel engine evaluated the effects on brake power (BP), brake specific fuel consumption (BSFC), brake thermal efficiency (BTE), brake mean effective pressure (BMEP), and volumetric efficiency (VE) across various load conditions (2.5 %, 25 %, 50 %, 75 %, and 100 % load). Additionally, a novel predictive model was developed using the Partial Reinforcement Optimiser (PRO) algorithm integrated with Random Vector Functional Link (RVFL) to enhance engine performance estimation. Comparative analysis with established optimisation algorithms (GWO, WOA, AOA, HHO, and traditional RVFL) demonstrated the superior accuracy of the PRO-RVFL model. The model consistently achieved the highest R2 and lowest RMSE scores for all evaluated parameters (BP: R2 ≈ 93 %, RMSE ≈ 1.13; BSFC: R2 ≈ 91 %, RMSE ≈ 1.45; BTE: R2 ≈ 89 %; BMEP: R2 ≈ 81 %, RMSE ≈ 2.80; VE: R2 ≈ 71 %, RMSE ≈ 3.13). The findings support the viability of methanol/diesel blends in enhancing engine performance while promoting sustainability in transportation. This study, with its precise experimentation and advanced modelling techniques, paves the way for the development of cleaner and more efficient transportation systems. [ABSTRACT FROM AUTHOR]

The recent advances in artificial intelligence modern approaches can play vital roles in the Internet of Medical Things (IoMT). Automatic diagnosis is one of the most important topics in the IoMT, including cancer diagnosis. Breast cancer is one of the top causes of death among women. Accurate diagnosis and early detection of breast cancer can improve the survival rate of patients. Deep learning models have demonstrated outstanding potential in accurately detecting and diagnosing breast cancer. This paper proposes a novel technology for breast cancer detection using CrossViT as the deep learning model and an enhanced version of the Growth Optimizer algorithm (MGO) as the feature selection method. CrossVit is a hybrid deep learning model that combines the strengths of both convolutional neural networks (CNNs) and transformers. The MGO is a meta-heuristic algorithm that selects the most relevant features from a large pool of features to enhance the performance of the model. The developed approach was evaluated on three publicly available breast cancer datasets and achieved competitive performance compared to other state-of-the-art methods. The results show that the combination of CrossViT and the MGO can effectively identify the most informative features for breast cancer detection, potentially assisting clinicians in making accurate diagnoses and improving patient outcomes. The MGO algorithm improves accuracy by approximately 1.59% on INbreast, 5.00% on MIAS, and 0.79% on MiniDDSM compared to other methods on each respective dataset. The developed approach can also be utilized to improve the Quality of Service (QoS) in the healthcare system as a deployable IoT-based intelligent solution or a decision-making assistance service, enhancing the efficiency and precision of the diagnosis. [Display omitted] • Develop a new breast cancer detection method using CrossViT and a modified GOA. • Apply CrossViT to extract features and to learn new representations from images. • Enhance feature selection using modified GOA based on Adaptive Gaussian distribution. • Apply proposed technique to enhance breast cancer detection using different datasets. [ABSTRACT FROM AUTHOR]

Attention deficit hyperactivity disorder (ADHD) has high incidence rate among children which may be due to excessive monosodium glutamate (MSG) consumption and social isolation (SI). We aimed to explore the relationships between MSG, SI, and ADHD development and to evaluate the neuroprotective potential of Punicalagin (PUN). Eighty male rat pups randomly distributed into eight groups. Group I is the control, and Group II is socially engaged rats treated with PUN. Groups III to VII were exposed to ADHD-inducing factors: Group III to SI, Group IV to MSG, and Group V to both SI and MSG. Furthermore, Groups VI to VIII were the same Groups III to V but additionally received PUN treatment. Exposure to MSG and/or SI led to pronounced behavioral anomalies, histological changes and indicative of ADHD-like symptoms in rat pups which is accompanied by inhibition of the nuclear factor erythroid 2-related factor 2 (Nrf2)/Heme-oxygenase 1 (HO-1)/Glutathione (GSH) pathway, decline of the brain-derived neurotrophic factor (BDNF) expression and activation of the Toll-like receptor 4 (TLR4)/Nuclear factor kappa B (NF-kB)/NLR Family Pyrin Domain Containing 3 (NLRP3) pathway. This resulted in elevated inflammatory biomarker levels, neuronal apoptosis, and disrupted neurotransmitter equilibrium. Meanwhile, pretreatment with PUN protected against all the previous alterations. We established compelling associations between MSG consumption, SI, and ADHD progression. Moreover, we proved that PUN is a promising neuroprotective agent against all risk factors of ADHD. [Display omitted] [ABSTRACT FROM AUTHOR]

The Internet of Things (IoT) enabled technology will be adopted to develop smart cities, electronic commerce, electronic learning, electronic health, and other aspects of online activities. IoT enabled pervasive and wide connectivity to many objects and services. Therefore, it is easy to target IoT and cloud malware infection. Thus, cybersecurity is an essential problem to build robust IoT systems. This paper leverages the recent developments of the swarm intelligence (SI) algorithms combined with the advances of deep neural networks to build an efficient intrusion detection system for IoT-cloud based environments. First, deep neural networks are used to obtain optimal features from the IoT IDS data. Then, an efficient feature selection technique is proposed based on a recently developed SI optimizer called Capuchin Search Algorithm (CapSA). The performance of the developed model, called CNN-CapSA, is tested with four IoT-Cloud datasets, namely, NSL-KDD, BoT-IoT, KDD99, and CIC2017. Moreover, we consider extensive empirical comparisons to other optimization algorithms using several classification performance measures. The outcomes verified that the developed approach has a competitive performance overall datasets. • A new IDS technique is proposed based on deep learning and swarm intelligence. • Propose a light feature extraction approach using The CNN. • Propose an efficient feature selection method based on CapSA optimizer. • Implement extensive evaluation comparisons using different public datasets. [ABSTRACT FROM AUTHOR]

In this paper, an improved Barnacles Mating Optimizer (BMO) is proposed to deal with optimization problems and develop a new automatic clustering approach. BMO is a well-established optimization technique inspired by the mating behavior of barnacles in real-life. The exploratory trends of BMO are influential and can maintain the right balance among exploration and exploitation. However, this population-based method can be improved further to reduce the probability of potential drawbacks for any optimization technique. As such, we revised the core searching phased of BMO based on a sine–cosine algorithm (SCA) and disruption operators (DO). The proposed method is named BMSCD, which updates the current solution by switching between the mechanisms of the BMO and SCA based on a probability calculated using the fitness value of the current solution. The experiments results on various benchmark cases for global optimizations demonstrate the improved performance of the proposed BMSCD in terms of quality of solutions, the balance of the exploration–exploitation, and convergence rates. Besides, the proposed BMSCD is evaluated by nine measures in solving different clustering problems. The results show that the BMSCD can effectively and powerfully address the tested problems and provide excellent performance compared to the state-of-the-art methods. • Developed an improved Sine–Cosine-Barnacles Algorithm Optimizer. • Evaluate the proposed method against benchmark functions and data clustering problems. • Compared the proposed method to other well-known methods. • Demonstrated effectiveness and superiority of the proposed method. [ABSTRACT FROM AUTHOR]

Ulcerative colitis (UC) is a chronic inflammatory condition that until this date, lacks curative treatments. Previously, synthetic selective CB2 receptor (CB2R) agonists demonstrated effective preclinical anti-inflammatory activities in UC. Phycocyanin (PC), photosynthetic assistant protein isolated from Microcystis aeruginosa Kützing blue green algae, has multiple pharmacological effects, however, it's effect against UC remains unexplored. Our study aimed at investigating the therapeutic effectiveness of PC against UC, and correlating its mechanisms with CB2R agonistic activities. In silico ; PC showed structural similarity with endocannabinoid receptors' ligand "Δ9-tetrahydrocannabinol", target prediction studies suggested high affinity for G-coupled protein family-receptors, and molecular docking affirmed preferable affinity towards CB2R vs CB1R. In LPS-exposed-Caco-2 cell line; PC demonstrated comparable interaction with CB2R, and downregulation of CB2R, p38 and MK2 gene expressions with reference agonist "6d", and exhibited preferred selectivity towards CB2R over CB1R. In DSS-induced mice; PC-treatment ameliorated DSS-induced colon shortening, elevated disease activity index, and colonic pathological alterations. PC showed effective CB2R activation through potent anti-inflammatory activities, Treg-cell accumulation, suppression in p38MAPK/MK2 signaling, and tight junction barrier restoration as indicated by ultrastructural examinations, elevated ZO-1 and occludin protein expressions, and Ki67 immunohistochemical expression in colonic tissues. Additionally, PC alleviated intestinal dysbiosis via downregulating LPS/TLR4/NF-κB signaling and gut microbiota maintenance. Notably, PC-protective activities were abolished when co-administered with SR144528 (selective CB2 antagonist) except for gut microbiota maintenance, which was independent from CB2R activation. Our findings provide evidence of PC effectiveness against UC through acting as CB2R agonist, thus expanding its possible therapeutic application against other inflammatory diseases. [Display omitted] • Phycocyanin (PC) is isolated from Microcystis aeruginosa Kützing algae. • In silico ; PC showed structural similarity and high affinity with CB2 receptors. • In vitro; PC showed potent interaction with CB2 receptors as selective agonist "6d". • In UC; PC activated CB2 receptors via suppressing inflammatory responses & p38/MK2. • PC maintained intestinal homeostasis via inhibiting TLR4/NF-κB signaling. [ABSTRACT FROM AUTHOR]

The current study investigated the neuroprotective activity of some drugs and nutriceuticals with antioxidant and anti-inflammatory potential on the pathogenesis of Parkinson's disease (PD). Rats were categorized into seven groups: Rats received tween80 daily for 5 weeks as a control group, MnCl 2 (10 mg/kg, i.p) either alone (group II) or in combination with vinpocetine (VIN) (20 mg/kg) (group III), punicalagin (PUN) (30 mg/kg) (group IV), niacin (85 mg/kg) (group V), vitamin E (Vit E) (100 mg/kg) (group VI) or their combination (group VII). Motor activities was examined using open-field and catalepsy. Striatal monamines, acetylcholinesterase, excitatory/inhibitory neurotransmitters, redox status, pro-oxidant content, brain inflammatory, apoptotic and antioxidant biomarkers levels were assessed. Besides, histopathological investigations of different brain regions were determined. Groups (IV –GVII) showed improved motor functions of PD rats. Applied drugs significantly increased the brain levels of monoamines with the strongest effect to PUN. Meanwhile, they significantly decreased levels of acetylcholinesterase with a strongest effect to PUN. Moreover, they exhibited significant neuronal protection and anti-inflammatory abilities through significant reduction of the brain levels of COX2, TNF-α and Il-1β with a strongest effect to the PUN. Interestingly; groups (IV – GVII) showed restored glutamate/GABA balance and exhibited a pronounced decrease in caspase-3 content and GSK-3β protein expression levels. In addition, they significantly increased Bcl2 mRNA expression levels with a strongest effect for PUN. All these findings were further confirmed by the histopathological examinations. As a conclusion, we propose VIN and PUN to mitigate the progression of PD via their antioxidant, anti-inflammatory, anti-apoptotic, neurotrophic and neurogenic activities. [Display omitted] • MnCl 2 induces biochemical changes in the brain: reduction of monoamines and disturbing the imbalance between excitatory and inhibitory neurotransmitters. • MnCl 2 changes the redox status and the inflammatory biomarkers of the brain tissues. • Punicalagin has maximum protective against decreased BDNF, monoamines, antioxidant contents levels and increased level of pro-oxidant content, brain cytokines, inflammatory and apoptotic markers. • Vinpocetine has maximum protective against decreased AChE, GABA and antioxidant contents and Bcl2 levels and increased level of glutamate. [ABSTRACT FROM AUTHOR]

• Developed a global optimization approach using cooperative meta-heuristic methods. • The proposed method inspired from the natural selection theory. • DE, GWO, WOA, SSA, SCA, and SOS are used to build the proposed method. • Three variants of proposed are developed based on strategy of updating solutions. • Results of proposed method compared with other methods using CEC2014 and CEC2017. This paper presents an alternative global optimization meta-heuristics (MHs) approach, inspired by the natural selection theory. The proposed approach depends on the competition among six MHs that allows generating an offspring, which can breed the high characteristics of parents since they are unique and competitive. Therefore, this leads to improve the convergence of the solutions towards an optimal solution and also, to avoid the limitations of other methods that aim to balance between exploitation and exploration. The six algorithms are differential evolution, whale optimization algorithm, grey wolf optimization, symbiotic organisms search algorithm, sine–cosine algorithm, and salp swarm algorithm. According to these algorithms, three variants of the proposed method are developed, in the first variant, one of the six algorithms will be used to update the current individual based on a predefined order and the probability of the fitness function for each individual. Whereas, the second variant updates each individual by permuting the six algorithms, then using the algorithms in the current permutation to update individuals. The third variant is considered as an extension of the second variant, which updates all individuals using only one algorithm from the six algorithms. Three different experiments are carried out using CEC 2014 and CEC 2017 benchmark functions to evaluate the efficiency of the proposed approach. Moreover, the proposed approach is compared with well known MH methods, including the six methods used to build it. Comparison results confirmed the efficiency of the proposed approach compared to other approaches according to different performance measures. [ABSTRACT FROM AUTHOR]