Monarch butterfly optimization

Multi objective elk herd optimization for efficient structural design

This research presents an advancement of the Elk Herd Optimization targeting specific real-world multi-objective optimization problems, this algorithm is stated as the multi-objective Elk Herd Optimization (MOEHO). MOEHO exploits reproductive behaviour among elk herds for balancing exploration and exploitation within the optimization procedure toward diversification and convergence. The algorithm performed better over the set of small-to-medium scale structural design problems thus is widely applicable in engineering design. Further, when compared with eight benchmark truss structures against five well-established algorithms the MOEHO has outperformed them in the perspective of performance parameters like Spacing (SP), Hypervolume (HV) and Inverted Generational Distance (IGD). More concrete statistical analysis through Friedman rank test also ascertains the robustness and efficiency of the algorithm, especially at high complexities in optimization. The research attracts attention to the ability of such an algorithm which maintains a balance between the exploration and exploitation. Computational efficiency of MOEHO and qualitatively diversifying solutions along Pareto front, makes it especially applicable in complex engineering applications. Further research into extension of MOEHO with applicability on more dimensional problems, applied even in energy systems optimization.

AI optimization algorithms enhance higher education management and personalized teaching through empirical analysis

This research investigates the application of artificial intelligence (AI) optimization algorithms in higher education management and personalized teaching. Through a comprehensive literature review, theoretical analysis, and empirical study, the potential, effectiveness, and challenges of integrating AI algorithms into educational processes and systems are explored. The study demonstrates that AI optimization algorithms can effectively solve complex educational management problems and enable personalized learning experiences. An empirical study conducted over one academic semester shows significant improvements in students' learning outcomes, engagement, satisfaction, and efficiency when using AI-driven personalized teaching compared to traditional approaches. The research also identifies challenges and limitations, including data privacy issues, algorithmic bias, and the need for human-AI interaction. Recommendations for future research directions are provided, emphasizing the importance of developing more adaptive algorithms, investigating long-term effects, and establishing ethical frameworks for AI in education.

Multi-resource constrained elective surgical scheduling with Nash equilibrium toward smart hospitals

This paper focuses on the elective surgical scheduling problem with multi-resource constraints, including material resources, such as operating rooms (ORs) and non-operating room (NOR) beds, and human resources (i.e., surgeons, anesthesiologists, and nurses). The objective of multi-resource constrained elective surgical scheduling (MESS) is to simultaneously minimize the average recovery completion time for all patients, the average overtime for medical staffs, and the total medical cost. This problem can be formulated as a mixed integer linear multi-objective optimization model, and the honey badger algorithm based on the Nash equilibrium (HBA-NE) is developed for the MESS. Experimental studies were carried out to test the performance of the proposed approach, and the performance of the proposed surgical scheduling scheme was validated. Finally, to narrow the gap between the optimal surgical scheduling solution and actual hospital operations, digital twin (DT) technology is adopted to build a physical-virtual hospital surgery simulation model. The experimental results show that by introducing a digital twin, the physical and virtual spaces of the smart hospital can be integrated to visually simulate and verify surgical processes.

Optimized Hybrid Deep Learning Framework for Early Detection of Alzheimer's Disease Using Adaptive Weight Selection

BACKGROUND

Alzheimer's disease (AD) is a progressive neurological disorder that significantly affects middle-aged and elderly adults, leading to cognitive deterioration and hindering daily activities. Notwithstanding progress, conventional diagnostic techniques continue to be susceptible to inaccuracies and inefficiencies. Timely and precise diagnosis is essential for early intervention.

METHODS

We present an enhanced hybrid deep learning framework that amalgamates the EfficientNetV2B3 with Inception-ResNetV2 models. The models were integrated using an adaptive weight selection process informed by the Cuckoo Search optimization algorithm. The procedure commences with the pre-processing of neuroimaging data to guarantee quality and uniformity. Features are subsequently retrieved from the neuroimaging data by utilizing the EfficientNetV2B3 and Inception-ResNetV2 models. The Cuckoo Search algorithm allocates weights to various models dynamically, contingent upon their efficacy in particular diagnostic tasks. The framework achieves balanced usage of the distinct characteristics of both models through the iterative optimization of the weight configuration. This method improves classification accuracy, especially for early-stage Alzheimer's disease. A thorough assessment was conducted on extensive neuroimaging datasets to verify the framework's efficacy.

RESULTS

The framework attained a Scott's Pi agreement score of 0.9907, indicating exceptional diagnostic accuracy and dependability, especially in identifying the early stages of Alzheimer's disease. The results show its superiority over current state-of-the-art techniques.

CONCLUSIONS

The results indicate the substantial potential of the proposed framework as a reliable and scalable instrument for the identification of Alzheimer's disease. This method effectively mitigates the shortcomings of conventional diagnostic techniques and current deep learning algorithms by utilizing the complementing capabilities of EfficientNetV2B3 and Inception-ResNetV2 by using an optimized weight selection mechanism. The adaptive characteristics of the Cuckoo Search optimization facilitate its application across many diagnostic circumstances, hence extending its utility to a wider array of neuroimaging datasets. The capacity to accurately identify early-stage Alzheimer's disease is essential for facilitating prompt therapies, which are crucial for decelerating disease development and enhancing patient outcomes.

MobileNet-V2 /IFHO model for Accurate Detection of early-stage diabetic retinopathy

Diabetic retinopathy is a serious eye disease that may lead to loss of vision if it is not treated. Early detection is crucial in preventing further vision impairment and enabling timely interventions. Despite notable advancements in AI-based methods for detecting diabetic retinopathy, researchers are still striving to enhance the efficiency of these techniques. Therefore, obtaining an efficient technique in this field is essential. In this research, a new strategy has been proposed to improve the detection of diabetic retinopathy by increasing the accuracy of diagnosis and identifying cases in the initial stages. To achieve this, it has been proposed to integrate the MobileNet-V2 deep learning-based neural network with Improved Fire Hawk Optimizer (IFHO). The MobileNet-V2 network has been renowned for its efficiency and accuracy in image classification tasks, making it a suitable candidate for diabetic retinopathy detection. By combining it with the IFHO, the feature selection process has been optimized, which is essential for identifying relevant patterns and abnormalities related to diabetic retinopathy. The Diabetic Retinopathy 2015 dataset has been used to evaluate the effectiveness of the MobileNet-V2/IFHO model. The study results indicate that the DRMNV2/IFHO model consistently outperforms other methods in terms of precision, accuracy, and recall. Specifically, the model achieves an average precision of 97.521 %, accuracy of 96.986 %, and recall of 98.543 %. Moreover, when compared to advanced techniques, the DRMNV2/IFHO model demonstrates superior performance in specificity, F1-score, and AUC, with average values of 97.233 %, 93.8 %, and 0.927, respectively. These results underscore the potential of the DRMNV2/IFHO model as a valuable tool for improving the accuracy and efficiency of DR diagnosis. Nevertheless, additional validation and testing on larger datasets are required to verify the model's effectiveness and robustness in real-world clinical scenarios.

Taylor Remora optimization enabled deep learning algorithms for percentage of pesticide detection in grapes

In the world, grapes are considered as the most significant fruit, and it comprises various nutrients, like Vitamin C and it is utilized to produce wines and raisins. The major six general grape leaf diseases and pests are brown spots, leaf blight, downy mildew, anthracnose, and black rot. However, the existing manual detection methods are time-consuming and require more efforts. In this paper, an effectual grape leaf disease finding and percentage of pesticide detection approach is devised usingan optimized deep learning scheme. Here, the input image is pre-processed and then, black spot segmentation is done using proposed Taylor Remora Optimization Procedure (TROA). The TROA is the combination of Taylor concept and Remora Optimization Algorithm (ROA). After that, the multi-classification of grape leaf disease is performed to classify the disease as Black rot, Black measles, Isariopsis leaf spot and healthy. Accordingly, the training process of the Deep Neuro-Fuzzy Optimizer (DNFN) is done using Sine Cosine Butterfly Optimization (SCBO). Then, pesticide classification is done using Deep Maxout Network (DMN) and the training of the DMN is done using the Monarch Anti Corona Optimization (MACO) algorithm. Finally, the pesticide percentage level detection is performed using Deep Belief Network (DBN), which is trained by the TROA. The devised scheme obtained highest accuracy of 0.9327, sensitivity of 0.9383, and 0.9429. Thus, this method can assist as an effectual decision provision system for assisting the farmers to find the percentage of pesticide affected in grape leaf diseases.

Nature-Inspired Intelligent Computing: A Comprehensive Survey

Nature, with its numerous surprising rules, serves as a rich source of creativity for the development of artificial intelligence, inspiring researchers to create several nature-inspired intelligent computing paradigms based on natural mechanisms. Over the past decades, these paradigms have revealed effective and flexible solutions to practical and complex problems. This paper summarizes the natural mechanisms of diverse advanced nature-inspired intelligent computing paradigms, which provide valuable lessons for building general-purpose machines capable of adapting to the environment autonomously. According to the natural mechanisms, we classify nature-inspired intelligent computing paradigms into 4 types: evolutionary-based, biological-based, social-cultural-based, and science-based. Moreover, this paper also illustrates the interrelationship between these paradigms and natural mechanisms, as well as their real-world applications, offering a comprehensive algorithmic foundation for mitigating unreasonable metaphors. Finally, based on the detailed analysis of natural mechanisms, the challenges of current nature-inspired paradigms and promising future research directions are presented.

Mapping biomimicry research to sustainable development goals

This study systematically evaluates biomimicry research within the context of sustainable development goals (SDGs) to discern the interdisciplinary interplay between biomimicry and SDGs. The alignment of biomimicry with key SDGs showcases its interdisciplinary nature and potential to offer solutions across the health, sustainability, and energy sectors. This study identified two primary thematic clusters. The first thematic cluster focused on health, partnership, and life on land (SDGs 3, 17, and 15), highlighting biomimicry's role in healthcare innovations, sustainable collaboration, and land management. This cluster demonstrates the potential of biomimicry to contribute to medical technologies, emphasizing the need for cross-sectoral partnerships and ecosystem preservation. The second thematic cluster revolves around clean water, energy, infrastructure, and marine life (SDGs 6, 7, 9, and 14), showcasing nature-inspired solutions for sustainable development challenges, including energy generation and water purification. The prominence of SDG 7 within this cluster indicates that biomimicry significantly contributes to sustainable energy practices. The analysis of thematic clusters further revealed the broad applicability of biomimicry and its role in enhancing sustainable energy access and promoting ecosystem conservation. Emerging research topics, such as metaheuristics, nanogenerators, exosomes, and bioprinting, indicate a dynamic field poised for significant advancements. By mapping the connections between biomimicry and SDGs, this study provides a comprehensive overview of the field's trajectory, emphasizing its importance in advancing global sustainability efforts.

Utilizing sine trigonometric q-spherical fuzzy rough aggregation operators for group decision-making and their role in digital transformation

q-spherical fuzzy rough set (q-SFRS) is also one of the fundamental concepts for addressing more uncertainties in decision problems than the existing structures of fuzzy sets, and thus its implementation was more substantial. The well-known sine trigonometric function maintains the periodicity and symmetry of the origin in nature and thus satisfies the expectations of the experts over the multi-parameters. Taking this feature and the significance of the q-SFRSs into consideration, the main objective of the article is to describe some reliable sine trigonometric laws for SFSs. Associated with these laws, we develop new average and geometric aggregation operators to aggregate the q-spherical fuzzy rough numbers. Then, we presented a group decision-making strategy to address the multi-attribute group decision-making problem using the developed aggregation operators. To verify the value of the defined operators, a MAGDM strategy is provided along with applications for selecting a Cloud Service Provider and a Digital Transformation Vendor for digital transformation. Moreover, a comparative study is also performed to present the effectiveness of the developed approach.

A novel multi-strategy ameliorated quasi-oppositional chaotic tunicate swarm algorithm for global optimization and constrained engineering applications

Over the last few decades, a number of prominent meta-heuristic algorithms have been put forth to address complex optimization problems. However, there is a critical need to enhance these existing meta-heuristics by employing a variety of evolutionary techniques to tackle the emerging challenges in engineering applications. As a result, this study attempts to boost the efficiency of the recently introduced bio-inspired algorithm, the Tunicate Swarm Algorithm (TSA), which is motivated by the foraging and swarming behaviour of bioluminescent tunicates residing in the deep sea. Like other algorithms, the TSA has certain limitations, including getting trapped in the local optimal values and a lack of exploration ability, resulting in premature convergence when dealing with highly challenging optimization problems. To overcome these shortcomings, a novel multi-strategy ameliorated TSA, termed the Quasi-Oppositional Chaotic TSA (QOCTSA), has been proposed as an enhanced variant of TSA. This enhanced method contributes the simultaneous incorporation of the Quasi-Oppositional Based Learning (QOBL) and Chaotic Local Search (CLS) mechanisms to effectively balance exploration and exploitation. The implementation of QOBL improves convergence accuracy and exploration rate, while the inclusion of a CLS strategy with ten chaotic maps improves exploitation by enhancing local search ability around the most prospective regions. Thus, the QOCTSA significantly enhances convergence accuracy while maintaining TSA diversification. The experimentations are conducted on a set of thirty-three diverse functions: CEC2005 and CEC2019 test functions, as well as several real-world engineering problems. The statistical and graphical outcomes indicate that QOCTSA is superior to TSA and exhibits a faster rate of convergence. Furthermore, the statistical tests, specifically the Wilcoxon rank-sum test and t-test, reveal that the QOCTSA method outperforms the other competing algorithms in the domain of real-world engineering design problems.

A Multi-Objective Optimization Problem Solving Method Based on Improved Golden Jackal Optimization Algorithm and Its Application

The traditional golden jackal optimization algorithm (GJO) has slow convergence speed, insufficient accuracy, and weakened optimization ability in the process of finding the optimal solution. At the same time, it is easy to fall into local extremes and other limitations. In this paper, a novel golden jackal optimization algorithm (SCMGJO) combining sine-cosine and Cauchy mutation is proposed. On one hand, tent mapping reverse learning is introduced in population initialization, and sine and cosine strategies are introduced in the update of prey positions, which enhances the global exploration ability of the algorithm. On the other hand, the introduction of Cauchy mutation for perturbation and update of the optimal solution effectively improves the algorithm's ability to obtain the optimal solution. Through the optimization experiment of 23 benchmark test functions, the results show that the SCMGJO algorithm performs well in convergence speed and accuracy. In addition, the stretching/compression spring design problem, three-bar truss design problem, and unmanned aerial vehicle path planning problem are introduced for verification. The experimental results prove that the SCMGJO algorithm has superior performance compared with other intelligent optimization algorithms and verify its application ability in engineering applications.

FIELD TRAVERSAL PATH PLANNING FOR AGRICULTURAL ROBOTS IN HILLY AREAS BASED ON DISCRETE ARTIFICIAL BEE COLONY ALGORITHM

In this study, the discrete artificial bee colony (DABC) algorithm was proposed to plan the path of agricultural robots traversing multiple fields in hilly areas. Based on the basic ABC algorithm as the framework, the path coding method was adopted, and the discrete crossover operator, reverse operator, immune operator, and single/multi-step 2-opt operator were comprehensively used to help hired bees, observing bees, and scout bees to generate new food sources. Finally, the optimized field traversal order and the entrance and exit distribution of each field were obtained. The simulation results showed that compared with the traditional ABC algorithm, the average shortest path of the DABC algorithm proposed in this study was shortened by 1.59%, accompanied by the less iterations contributing to algorithm convergence and good ability to jump out of the local optimal solution. The simulation experiment was carried out using real field data and field operation parameters. The field traversal order and the entrance and exit distribution obtained by the proposed method can effectively reduce the length of the transfer path and its repeatability. This study exhibits superiority and feasibility in the field traversal path planning of agricultural robots in hilly areas, and the trajectory coordinates output by the algorithm can provide a path reference for large-area operations of agricultural machinery drivers or unmanned agricultural machineries.

Archimedes Optimization Algorithm-Based Feature Selection with Hybrid Deep-Learning-Based Churn Prediction in Telecom Industries

Customer churn prediction (CCP) implies the deployment of data analytics and machine learning (ML) tools to forecast the churning customers, i.e., probable customers who may remove their subscriptions, thus allowing the companies to apply targeted customer retention approaches and reduce the customer attrition rate. This predictive methodology improves active customer management and provides enriched satisfaction to the customers and also continuous business profits. By recognizing and prioritizing the relevant features, such as usage patterns and customer collaborations, and also by leveraging the capability of deep learning (DL) algorithms, the telecom companies can develop highly robust predictive models that can efficiently anticipate and mitigate customer churn by boosting retention approaches. In this background, the current study presents the Archimedes optimization algorithm-based feature selection with a hybrid deep-learning-based churn prediction (AOAFS-HDLCP) technique for telecom companies. In order to mitigate high-dimensionality problems, the AOAFS-HDLCP technique involves the AOAFS approach to optimally choose a set of features. In addition to this, the convolutional neural network with autoencoder (CNN-AE) model is also involved for the churn prediction process. Finally, the thermal equilibrium optimization (TEO) technique is employed for hyperparameter selection of the CNN-AE algorithm, which, in turn, helps in achieving improved classification performance. A widespread experimental analysis was conducted to illustrate the enhanced performance of the AOAFS-HDLCP algorithm. The experimental outcomes portray the high efficiency of the AOAFS-HDLCP approach over other techniques, with a maximum accuracy of 94.65%.

Giant Armadillo Optimization: A New Bio-Inspired Metaheuristic Algorithm for Solving Optimization Problems

In this paper, a new bio-inspired metaheuristic algorithm called Giant Armadillo Optimization (GAO) is introduced, which imitates the natural behavior of giant armadillo in the wild. The fundamental inspiration in the design of GAO is derived from the hunting strategy of giant armadillos in moving towards prey positions and digging termite mounds. The theory of GAO is expressed and mathematically modeled in two phases: (i) exploration based on simulating the movement of giant armadillos towards termite mounds, and (ii) exploitation based on simulating giant armadillos' digging skills in order to prey on and rip open termite mounds. The performance of GAO in handling optimization tasks is evaluated in order to solve the CEC 2017 test suite for problem dimensions equal to 10, 30, 50, and 100. The optimization results show that GAO is able to achieve effective solutions for optimization problems by benefiting from its high abilities in exploration, exploitation, and balancing them during the search process. The quality of the results obtained from GAO is compared with the performance of twelve well-known metaheuristic algorithms. The simulation results show that GAO presents superior performance compared to competitor algorithms by providing better results for most of the benchmark functions. The statistical analysis of the Wilcoxon rank sum test confirms that GAO has a significant statistical superiority over competitor algorithms. The implementation of GAO on the CEC 2011 test suite and four engineering design problems show that the proposed approach has effective performance in dealing with real-world applications.

FedEHR: A Federated Learning Approach towards the Prediction of Heart Diseases in IoT-Based Electronic Health Records

In contemporary healthcare, the prediction and identification of cardiac diseases is crucial. By leveraging the capabilities of Internet of Things (IoT)-enabled devices and Electronic Health Records (EHRs), the healthcare sector can largely benefit to improve patient outcomes by increasing the accuracy of disease prediction. However, protecting data privacy is essential to promote participation and adhere to rules. The suggested methodology combines EHRs with IoT-generated health data to predict heart disease. For its capacity to manage high-dimensional data and choose pertinent features, a soft-margin L1-regularised Support Vector Machine (sSVM) classifier is used. The large-scale sSVM problem is successfully solved using the cluster primal-dual splitting algorithm, which improves computational complexity and scalability. The integration of federated learning provides a cooperative predictive analytics methodology that upholds data privacy. The use of a federated learning framework in this study, with a focus on peer-to-peer applications, is crucial for enabling collaborative predictive modeling while protecting the confidentiality of each participant's private medical information.

OOBO: A New Metaheuristic Algorithm for Solving Optimization Problems

This study proposes the One-to-One-Based Optimizer (OOBO), a new optimization technique for solving optimization problems in various scientific areas. The key idea in designing the suggested OOBO is to effectively use the knowledge of all members in the process of updating the algorithm population while preventing the algorithm from relying on specific members of the population. We use a one-to-one correspondence between the two sets of population members and the members selected as guides to increase the involvement of all population members in the update process. Each population member is chosen just once as a guide and is only utilized to update another member of the population in this one-to-one interaction. The proposed OOBO's performance in optimization is evaluated with fifty-two objective functions, encompassing unimodal, high-dimensional multimodal, and fixed-dimensional multimodal types, and the CEC 2017 test suite. The optimization results highlight the remarkable capacity of OOBO to strike a balance between exploration and exploitation within the problem-solving space during the search process. The quality of the optimization results achieved using the proposed OOBO is evaluated by comparing them to eight well-known algorithms. The simulation findings show that OOBO outperforms the other algorithms in addressing optimization problems and can give more acceptable quasi-optimal solutions. Also, the implementation of OOBO in six engineering problems shows the effectiveness of the proposed approach in solving real-world optimization applications.

A hierarchical multi-leadership sine cosine algorithm to dissolving global optimization and data classification: The COVID-19 case study

The Sine Cosine Algorithm (SCA) is an outstanding optimizer that is appreciably used to dissolve complicated real-world problems. Nevertheless, this algorithm lacks sufficient population diversification and a sufficient balance between exploration and exploitation. So, effective techniques are required to tackle the SCA's fundamental shortcomings. Accordingly, the present paper suggests an improved version of SCA called Hierarchical Multi-Leadership SCA (HMLSCA) which uses an effective hierarchical multi-leadership search mechanism to lead the search process on multiple paths. The efficiency of the HMLSCA has been appraised and compared with a set of famous metaheuristic algorithms to dissolve the classical eighteen benchmark functions and thirty CEC 2017 test suites. The results demonstrate that the HMLSCA outperforms all compared algorithms and that the proposed algorithm provided a promising efficiency. Moreover, the HMLSCA was applied to handle the medicine data classification by optimizing the support vector machine's (SVM) parameters and feature weighting in eight datasets. The experiential outcomes verify the productivity of the HMLSCA with the highest classification accuracy and a gain scoring 1.00 Friedman mean rank versus the other evaluated metaheuristic algorithms. Furthermore, the proposed algorithm was used to diagnose COVID-19, in which it attained the topmost accuracy of 98% in diagnosing the infection on the COVID-19 dataset, which proves the performance of the proposed search strategy.

Dynamic Coati Optimization Algorithm for Biomedical Classification Tasks

Medical datasets are primarily made up of numerous pointless and redundant elements in a collection of patient records. None of these characteristics are necessary for a medical decision-making process. Conversely, a large amount of data leads to increased dimensionality and decreased classifier performance in terms of machine learning. Numerous approaches have recently been put out to address this issue, and the results indicate that feature selection can be a successful remedy. To meet the various needs of input patterns, medical diagnostic tasks typically involve learning a suitable categorization model. The k-Nearest Neighbors algorithm (kNN) classifier's classification performance is typically decreased by the input variables' abundance of irrelevant features. To simplify the kNN classifier, essential attributes of the input variables have been searched using the feature selection approach. This paper presents the Coati Optimization Algorithm (DCOA) in a dynamic form as a feature selection technique where each iteration of the optimization process involves the introduction of a different feature. We enhance the exploration and exploitation capability of DCOA by employing dynamic opposing candidate solutions. The most impressive feature of DCOA is that it does not require any preparatory parameter fine-tuning to the most popular metaheuristic algorithms. The CEC'22 test suite and nine medical datasets with various dimension sizes were used to evaluate the performance of the original COA and the proposed dynamic version. The statistical results were validated using the Bonferroni-Dunn test and Kendall's W test and showed the superiority of DCOA over seven well-known metaheuristic algorithms with an overall accuracy of 89.7%, a feature selection of 24%, a sensitivity of 93.35% a specificity of 96.81%, and a precision of 93.90%.

Boosted federated learning based on improved Particle Swarm Optimization for healthcare IoT devices

As healthcare data becomes increasingly available from various sources, including clinical institutions, patients, insurance companies, and pharmaceutical industries, machine learning (ML) services are becoming more significant in healthcare-facing domains. Therefore, it is imperative to ensure the integrity and reliability of ML models to maintain the quality of healthcare services. Particularly due to the growing need for privacy and security, healthcare data has resulted in each Internet of Things (IoT) device being treated as an independent source of data, isolated from other devices. Moreover, the limited computational and communication capabilities of wearable healthcare devices hinder the applicability of traditional ML. Federated Learning (FL) is a paradigm that maintains data privacy by storing only learned models on a server and advances with data from scattered clients, making it ideal for healthcare applications where patient data must be safeguarded. The potential of FL to transform healthcare is significant, as it can enable the development of new ML-powered applications that can enhance the quality of care, lower costs, and improve patient outcomes. However, the accuracy of current Federated Learning aggregation methods suffers greatly in unstable network situations due to the high volume of weights transmitted and received. To address this issue, we propose an alternative approach to Federated Average (FedAvg) that updates the global model by gathering score values from learned models primarily utilized in Federated Learning, using an improved version of Particle Swarm Optimization (PSO) called FedImpPSO. This approach boosts the robustness of the algorithm in erratic network conditions. To further enhance the speed and efficiency of data exchange within a network, we modify the format of the data clients send to servers using the FedImpPSO method. The proposed approach is evaluated using the CIFAR-10 and CIFAR-100 datasets and a Convolutional Neural Network (CNN). We found that it yielded an average accuracy improvement of 8.14% over FedAvg and 2.5% over Federated PSO (FedPSO). This study evaluates the use of FedImpPSO in healthcare by training a deep-learning model over two case studies to evaluate the effectiveness of our approach in healthcare. The first case study involves the classification of COVID-19 using public datasets (Ultrasound and X-ray) and achieved an F1-measure of 77.90% and 92.16%, respectively. The second case study was conducted over the cardiovascular dataset, where our proposed FedImpPSO achieves 91.18% and 92% accuracy in predicting the existence of heart diseases. As a result, our approach demonstrates the effectiveness of using FedImpPSO to improve the accuracy and robustness of Federated Learning in unstable network conditions and has potential applications in healthcare and other domains where data privacy is critical.

Hybrid Slime Mold and Arithmetic Optimization Algorithm with Random Center Learning and Restart Mutation

The slime mold algorithm (SMA) and the arithmetic optimization algorithm (AOA) are two novel meta-heuristic optimization algorithms. Among them, the slime mold algorithm has a strong global search ability. Still, the oscillation effect in the later iteration stage is weak, making it difficult to find the optimal position in complex functions. The arithmetic optimization algorithm utilizes multiplication and division operators for position updates, which have strong randomness and good convergence ability. For the above, this paper integrates the two algorithms and adds a random central solution strategy, a mutation strategy, and a restart strategy. A hybrid slime mold and arithmetic optimization algorithm with random center learning and restart mutation (RCLSMAOA) is proposed. The improved algorithm retains the position update formula of the slime mold algorithm in the global exploration section. It replaces the convergence stage of the slime mold algorithm with the multiplication and division algorithm in the local exploitation stage. At the same time, the stochastic center learning strategy is adopted to improve the global search efficiency and the diversity of the algorithm population. In addition, the restart strategy and mutation strategy are also used to improve the convergence accuracy of the algorithm and enhance the later optimization ability. In comparison experiments, different kinds of test functions are used to test the specific performance of the improvement algorithm. We determine the final performance of the algorithm by analyzing experimental data and convergence images, using the Wilcoxon rank sum test and Friedman test. The experimental results show that the improvement algorithm, which combines the slime mold algorithm and arithmetic optimization algorithm, is effective. Finally, the specific performance of the improvement algorithm on practical engineering problems was evaluated.

Joint Light-Sensitive Balanced Butterfly Optimizer for Solving the NLO and NCO Problems of WSN for Environmental Monitoring

Existing swarm intelligence (SI) optimization algorithms applied to node localization optimization (NLO) and node coverage optimization (NCO) problems have low accuracy. In this study, a novel balanced butterfly optimizer (BBO) is proposed which comprehensively considers that butterflies in nature have both smell-sensitive and light-sensitive characteristics. These smell-sensitive and light-sensitive characteristics are used for the global and local search strategies of the proposed algorithm, respectively. Notably, the value of individuals' smell-sensitive characteristic is generally positive, which is a point that cannot be ignored. The performance of the proposed BBO is verified by twenty-three benchmark functions and compared to other state-of-the-art (SOTA) SI algorithms, including particle swarm optimization (PSO), differential evolution (DE), grey wolf optimizer (GWO), artificial butterfly optimization (ABO), butterfly optimization algorithm (BOA), Harris hawk optimization (HHO), and aquila optimizer (AO). The results demonstrate that the proposed BBO has better performance with the global search ability and strong stability. In addition, the BBO algorithm is used to address NLO and NCO problems in wireless sensor networks (WSNs) used in environmental monitoring, obtaining good results.

Multi-threshold image segmentation of 2D OTSU inland ships based on improved genetic algorithm

Waterway transportation is a crucial mode of transportation, but ensuring navigational safety in waterways requires effective guidance of ships by the Water Resources Bureau. However, supervisors may only be interested in the ship portion of a complex image and need to quickly obtain relevant ship information. Therefore, this paper proposes a two-dimensional OTSU inland ships multi-threshold image segmentation algorithm based on the improved genetic algorithm. The improved algorithm enhances search accuracy and efficiency, improving image thresholding accuracy and reducing algorithm time complexity. Experimental verification shows the algorithm has excellent evaluation indexes and can achieve real-time segmentation of complex images. This method can not only address the challenges of complex inland navigation environments and difficult acquisition of target data sets, but also be applied to optimization problems in other fields by combining various metaheuristic algorithms.

Golf Optimization Algorithm: A New Game-Based Metaheuristic Algorithm and Its Application to Energy Commitment Problem Considering Resilience

In this research article, we uphold the principles of the No Free Lunch theorem and employ it as a driving force to introduce an innovative game-based metaheuristic technique named Golf Optimization Algorithm (GOA). The GOA is meticulously structured with two distinctive phases, namely, exploration and exploitation, drawing inspiration from the strategic dynamics and player conduct observed in the sport of golf. Through comprehensive assessments encompassing fifty-two objective functions and four real-world engineering applications, the efficacy of the GOA is rigorously examined. The results of the optimization process reveal GOA's exceptional proficiency in both exploration and exploitation strategies, effectively striking a harmonious equilibrium between the two. Comparative analyses against ten competing algorithms demonstrate a clear and statistically significant superiority of the GOA across a spectrum of performance metrics. Furthermore, the successful application of the GOA to the intricate energy commitment problem, considering network resilience, underscores its prowess in addressing complex engineering challenges. For the convenience of the research community, we provide the MATLAB implementation codes for the proposed GOA methodology, ensuring accessibility and facilitating further exploration.

An improved particle swarm optimization combined with double-chaos search

Particle swarm optimization (PSO) has been successfully applied to various complex optimization problems due to its simplicity and efficiency. However, the update strategy of the standard PSO algorithm is to learn from the global best particle, making it difficult to maintain diversity in the population and prone to premature convergence due to being trapped in local optima. Chaos search mechanism is an optimization technique based on chaotic dynamics, which utilizes the randomness and nonlinearity of a chaotic system for global search and can escape from local optima. To overcome the limitations of PSO, an improved particle swarm optimization combined with double-chaos search (DCS-PSO) is proposed in this paper. In DCS-PSO, we first introduce double-chaos search mechanism to narrow the search space, which enables PSO to focus on the neighborhood of the optimal solution and reduces the probability that the swarm gets trapped into a local optimum. Second, to enhance the population diversity, the logistic map is employed to perform a global search in the narrowed search space and the best solution found by both the logistic and population search guides the population to converge. Experimental results show that DCS-PSO can effectively narrow the search space and has better convergence accuracy and speed in most cases.

Privacy-Aware Collaborative Learning for Skin Cancer Prediction

Cancer, including the highly dangerous melanoma, is marked by uncontrolled cell growth and the possibility of spreading to other parts of the body. However, the conventional approach to machine learning relies on centralized training data, posing challenges for data privacy in healthcare systems driven by artificial intelligence. The collection of data from diverse sensors leads to increased computing costs, while privacy restrictions make it challenging to employ traditional machine learning methods. Researchers are currently confronted with the formidable task of developing a skin cancer prediction technique that takes privacy concerns into account while simultaneously improving accuracy. In this work, we aimed to propose a decentralized privacy-aware learning mechanism to accurately predict melanoma skin cancer. In this research we analyzed federated learning from the skin cancer database. The results from the study showed that 92% accuracy was achieved by the proposed method, which was higher than baseline algorithms.

Effective detection of Alzheimer's disease by optimizing fuzzy K-nearest neighbors based on salp swarm algorithm

Alzheimer's disease (AD) is a typical senile degenerative disease that has received increasing attention worldwide. Many artificial intelligence methods have been used in the diagnosis of AD. In this paper, a fuzzy k-nearest neighbor method based on the improved binary salp swarm algorithm (IBSSA-FKNN) is proposed for the early diagnosis of AD, so as to distinguish between patients with mild cognitive impairment (MCI), Alzheimer's disease (AD) and normal controls (NC). First, the performance and feature selection accuracy of the method are validated on 5 different benchmark datasets. Secondly, the paper uses the Structural Magnetic Resolution Imaging (sMRI) dataset, in terms of classification accuracy, sensitivity, specificity, etc., the effectiveness of the method on the AD dataset is verified. The simulation results show that the classification accuracy of this method for AD and MCI, AD and NC, MCI and NC are 95.37%, 100%, and 93.95%, respectively. These accuracies are better than the other five comparison methods. The method proposed in this paper can learn better feature subsets from serial multimodal features, so as to improve the performance of early AD diagnosis. It has a good application prospect and will bring great convenience for clinicians to make better decisions in clinical diagnosis.

Modified reptile search algorithm with multi-hunting coordination strategy for global optimization problems

The reptile search algorithm (RSA) is a bionic algorithm proposed by Abualigah. et al. in 2020. RSA simulates the whole process of crocodiles encircling and catching prey. Specifically, the encircling stage includes high walking and belly walking, and the hunting stage includes hunting coordination and cooperation. However, in the middle and later stages of the iteration, most search agents will move towards the optimal solution. However, if the optimal solution falls into local optimum, the population will fall into stagnation. Therefore, RSA cannot converge when solving complex problems. To enable RSA to solve more problems, this paper proposes a multi-hunting coordination strategy by combining Lagrange interpolation and teaching-learning-based optimization (TLBO) algorithm's student stage. Multi-hunting cooperation strategy will make multiple search agents coordinate with each other. Compared with the hunting cooperation strategy in the original RSA, the multi-hunting cooperation strategy has been greatly improved RSA's global capability. Moreover, considering RSA's weak ability to jump out of the local optimum in the middle and later stages, this paper adds the Lens pposition-based learning (LOBL) and restart strategy. Based on the above strategy, a modified reptile search algorithm with a multi-hunting coordination strategy (MRSA) is proposed. To verify the above strategies' effectiveness for RSA, 23 benchmark and CEC2020 functions were used to test MRSA's performance. In addition, MRSA's solutions to six engineering problems reflected MRSA's engineering applicability. It can be seen from the experiment that MRSA has better performance in solving test functions and engineering problems.

A hybrid rough set shuffled frog leaping knowledge inference system for diagnosis of lung cancer disease

Abundant medical data are generated in the digital world every second. However, gathering helpful information from these data is difficult. Gathering useful information from the dataset is very advantageous and demanding. Besides, such data also contain many extraneous features that do not influence the foreboding accuracy while diagnosing a disease. The data must eliminate these extraneous features to get a better diagnosis. Ultimately, the minimized information system will lead to a better diagnosis. In this paper, we have introduced an incremental rough set shuffled frog leaping algorithm for knowledge inference. The proposed algorithm helps find minimum features from an information system while handling complex databases with uncertainty and incompleteness. The proposed rough set shuffled frog leaping knowledge inference model works in two phases. In the initial phase, the incremental rough set shuffled frog leaping algorithm is used to get the most relevant features. Identifying the relevant features is carried out using a fitness function, which uses the rough degree of dependency. The use of the fitness function identifies the much information with the minimum number of features. The purpose of feature selection is to identify a feature subset from an original set of features without reducing the predictive accuracy and to scale back the computation overhead in the data processing. In the second phase, a rough set is utilized for knowledge discovery in perception with rule generation. The selection of decision rules is carried out based on the accuracy of the decision rule and a predefined threshold value. An empirical analysis of the lung disease information system and a comparative study is conducted. Experimental outcomes exhibit that hybrid techniques express the feasibility of the proposed model while achieving better classification accuracy.

Multi-objective optimization determines when, which and how to fuse deep networks: An application to predict COVID-19 outcomes

The COVID-19 pandemic has caused millions of cases and deaths and the AI-related scientific community, after being involved with detecting COVID-19 signs in medical images, has been now directing the efforts towards the development of methods that can predict the progression of the disease. This task is multimodal by its very nature and, recently, baseline results achieved on the publicly available AIforCOVID dataset have shown that chest X-ray scans and clinical information are useful to identify patients at risk of severe outcomes. While deep learning has shown superior performance in several medical fields, in most of the cases it considers unimodal data only. In this respect, when, which and how to fuse the different modalities is an open challenge in multimodal deep learning. To cope with these three questions here we present a novel approach optimizing the setup of a multimodal end-to-end model. It exploits Pareto multi-objective optimization working with a performance metric and the diversity score of multiple candidate unimodal neural networks to be fused. We test our method on the AIforCOVID dataset, attaining state-of-the-art results, not only outperforming the baseline performance but also being robust to external validation. Moreover, exploiting XAI algorithms we figure out a hierarchy among the modalities and we extract the features' intra-modality importance, enriching the trust on the predictions made by the model.

Kohonen neural network and symbiotic-organism search algorithm for intrusion detection of network viruses

Introduction

The development of the Internet has made life much more convenient, but forms of network intrusion have become increasingly diversified and the threats to network security are becoming much more serious. Therefore, research into intrusion detection has become very important for network security.

Methods

In this paper, a clustering algorithm based on the symbiotic-organism search (SOS) algorithm and a Kohonen neural network is proposed.

Results

The clustering accuracy of the Kohonen neural network is improved by using the SOS algorithm to optimize the weights in the Kohonen neural network.

Discussion

Our approach was verified with the KDDCUP99 network intrusion data. The experimental results show that SOS-Kohonen can effectively detect intrusion. The detection rate was higher, and the false alarm rate was lower.

A novel particle swarm optimization based on hybrid-learning model

The convergence speed and the diversity of the population plays a critical role in the performance of particle swarm optimization (PSO). In order to balance the trade-off between exploration and exploitation, a novel particle swarm optimization based on the hybrid learning model (PSO-HLM) is proposed. In the early iteration stage, PSO-HLM updates the velocity of the particle based on the hybrid learning model, which can improve the convergence speed. At the end of the iteration, PSO-HLM employs a multi-pools fusion strategy to mutate the newly generated particles, which can expand the population diversity, thus avoid PSO-HLM falling into a local optima. In order to understand the strengths and weaknesses of PSO-HLM, several experiments are carried out on 30 benchmark functions. Experimental results show that the performance of PSO-HLM is better than other the-state-of-the-art algorithms.

Feature selection of pre-trained shallow CNN using the QLESCA optimizer: COVID-19 detection as a case study

According to the World Health Organization, millions of infections and a lot of deaths have been recorded worldwide since the emergence of the coronavirus disease (COVID-19). Since 2020, a lot of computer science researchers have used convolutional neural networks (CNNs) to develop interesting frameworks to detect this disease. However, poor feature extraction from the chest X-ray images and the high computational cost of the available models introduce difficulties for an accurate and fast COVID-19 detection framework. Moreover, poor feature extraction has caused the issue of 'the curse of dimensionality', which will negatively affect the performance of the model. Feature selection is typically considered as a preprocessing mechanism to find an optimal subset of features from a given set of all features in the data mining process. Thus, the major purpose of this study is to offer an accurate and efficient approach for extracting COVID-19 features from chest X-rays that is also less computationally expensive than earlier approaches. To achieve the specified goal, we design a mechanism for feature extraction based on shallow conventional neural network (SCNN) and used an effective method for selecting features by utilizing the newly developed optimization algorithm, Q-Learning Embedded Sine Cosine Algorithm (QLESCA). Support vector machines (SVMs) are used as a classifier. Five publicly available chest X-ray image datasets, consisting of 4848 COVID-19 images and 8669 non-COVID-19 images, are used to train and evaluate the proposed model. The performance of the QLESCA is evaluated against nine recent optimization algorithms. The proposed method is able to achieve the highest accuracy of 97.8086% while reducing the number of features from 100 to 38. Experiments prove that the accuracy of the model improves with the usage of the QLESCA as the dimensionality reduction technique by selecting relevant features.

Graphical abstract

Fast instruments and tissues segmentation of micro-neurosurgical scene using high correlative non-local network

Surgical scene segmentation provides critical information for guidance in micro-neurosurgery. Segmentation of instruments and critical tissues contributes further to robot assisted surgery and surgical evaluation. However, due to the lack of relevant scene segmentation dataset, scale variation and local similarity, micro-neurosurgical segmentation faces many challenges. To address these issues, a high correlative non-local network (HCNNet), is proposed to aggregate multi-scale feature by optimized non-local mechanism. HCNNet adopts two-branch design to generate features of different scale efficiently, while the two branches share common weights in shallow layers. Several short-term dense concatenate (STDC) modules are combined as the backbone to capture both semantic and spatial information. Besides, a high correlative non-local module (HCNM) is designed to guide the upsampling process of the high-level feature by modeling global context generated from the low-level feature. It filters out confused pixels of different classes in the non-local correlation map. Meanwhile, a large segmentation dataset named NeuroSeg is constructed, which contains 15 types of instruments and 3 types of tissues that appear in meningioma resection surgery. The proposed HCNNet achieves the state-of-the-art performance on NeuroSeg, it reaches an inference speed of 54.85 FPS with the highest accuracy of 59.62% mIoU, 74.7% Dice, 70.55% mAcc and 87.12% aAcc.

BrainTGL: A dynamic graph representation learning model for brain network analysis

Modeling the dynamics characteristics in functional brain networks (FBNs) is important for understanding the functional mechanism of the human brain. However, the current works do not fully consider the potential complex spatial and temporal correlations in human brain. To solve this problem, we propose a temporal graph representation learning framework for brain networks (BrainTGL). The framework involves a temporal graph pooling for eliminating the noisy edges as well as data inconsistency, and a dual temporal graph learning for capturing the spatio-temporal features of the temporal graphs. The proposed method has been evaluated in both tasks of brain disease (ASD, MDD and BD) diagnosis/gender classification (classification task) and subtype identification (clustering task) on the four datasets: Human Connectome Project (HCP), Autism Brain Imaging Data Exchange (ABIDE), NMU-MDD and NMU-BD. A large improvement is achieved for the ASD diagnosis. Specifically, our model outperforms the GroupINN and ST-GCN by an average increase of 4.2% and 8.6% on accuracy, respectively, demonstrating its advantages in comparison to the state-of-the-art methods based on functional connectivity features or learned spatio-temporal features. The results demonstrate that learning the spatial-temporal brain network representation for modeling dynamics characteristics in FBNs can improve the model's performance on both disease diagnosis and subtype identification tasks for multiple disorders. Apart from performance, the improvements of computational efficiency and convergence speed reduce training costs.

Optimization of pv cells/modules parameters using a modified quasi-oppositional logistic chaotic rao-1 (QOLCR) algorithm

Since the behavior of photovoltaic (PV) modules under different operational conditions is highly nonlinear, predicting the performance of PV systems in industrial applications is becoming a major challenge issue. Moreover, the most important information required to configure an optimal PV system is unavailable in all manufacturer's datasheets. In this context, a novel method is recommended to optimize PV cells/module parameters with the ability to correctly characterize the I-V and P-V curves of different PV models. In the present article, a chaotic map is incorporated in the so-called quasi-oppositional Rao-1 algorithm to improve its efficiency, and the resulting algorithm is named quasi-oppositional logistic chaotic Rao-1 (QOLCR) algorithm. Numerical results indicate that the QOLCR algorithm has presented very good performance in terms of accuracy and robustness. The idea is to minimize the root mean square error (RMSE) between the estimated and the actual data. Simulation results in the single diode model give an RMSE of value [Formula: see text], and in the double diode model, an RMSE of value [Formula: see text] has been reached as the minimum value among the other compared optimization methods. Hence, the QOLCR approach also converges faster than the basic Rao-1 algorithm and its other variants. Moreover, the modified QO Rao-1 algorithm shows its perfectness and could be involved as tools for optimal designing of PV systems.

HD2A-Net: A novel dual gated attention network using comprehensive hybrid dilated convolutions for medical image segmentation

The convolutional neural networks (CNNs) have been widely proposed in the medical image analysis tasks, especially in the image segmentations. In recent years, the encoder-decoder structures, such as the U-Net, were rendered. However, the multi-scale information transmission and effective modeling for long-range feature dependencies in these structures were not sufficiently considered. To improve the performance of the existing methods, we propose a novel hybrid dual dilated attention network (HD²A-Net) to conduct the lesion region segmentations. In the proposed network, we innovatively present the comprehensive hybrid dilated convolution (CHDC) module, which facilitates the transmission of the multi-scale information. Based on the CHDC module and the attention mechanisms, we design a novel dual dilated gated attention (DDGA) block to enhance the saliency of related regions from the multi-scale aspect. Besides, a dilated dense (DD) block is designed to expand the receptive fields. The ablation studies were performed to verify our proposed blocks. Besides, the interpretability of the HD²A-Net was analyzed through the visualization of the attention weight maps from the key blocks. Compared to the state-of-the-art methods including CA-Net, DeepLabV3+, and Attention U-Net, the HD²A-Net outperforms significantly, with the metrics of Dice, Average Symmetric Surface Distance (ASSD), and mean Intersection-over-Union (mIoU) reaching 93.16%, 93.63%, and 94.72%, 0.36 pix, 0.69 pix, and 0.52 pix, and 88.03%, 88.67%, and 90.33% on three publicly available medical image datasets: MAEDE-MAFTOUNI (COVID-19 CT), ISIC-2018 (Melanoma Dermoscopy), and Kvasir-SEG (Gastrointestinal Disease Polyp), respectively.

An efficient coverage method for SEMWSNs based on adaptive chaotic Gaussian variant snake optimization algorithm

Soil element monitoring wireless sensor networks (SEMWSNs) are widely used in soil element monitoring agricultural activities. SEMWSNs monitor changes in soil elemental content during agriculture products growing through nodes. Based on the feedback from the nodes, farmers adjust irrigation and fertilization strategies on time, thus promoting the economic growth of crops. The critical issue in SEMWSNs coverage studies is to achieve maximum coverage of the entire monitoring field by adopting a smaller number of sensor nodes. In this study, a unique adaptive chaotic Gaussian variant snake optimization algorithm (ACGSOA) is proposed for solving the above problem, which also has the advantages of solid robustness, low algorithmic complexity, and fast convergence. A new chaotic operator is proposed in this paper to optimize the position parameters of individuals, enhancing the convergence speed of the algorithm. Moreover, an adaptive Gaussian variant operator is also designed in this paper to effectively avoid SEMWSNs from falling into local optima during the deployment process. Simulation experiments are designed to compare ACGSOA with other widely used metaheuristics, namely snake optimizer (SO), whale optimization algorithm (WOA), artificial bee colony algorithm (ABC), and fruit fly optimization algorithm (FOA). The simulation results show that the performance of ACGSOA has been dramatically improved. On the one hand, ACGSOA outperforms other methods in terms of convergence speed, and on the other hand, the coverage rate is improved by 7.20%, 7.32%, 7.96%, and 11.03% compared with SO, WOA, ABC, and FOA, respectively.

Quantifying imbalanced classification methods for leukemia detection

Uncontrolled proliferation of B-lymphoblast cells is a common characterization of Acute Lymphoblastic Leukemia (ALL). B-lymphoblasts are found in large numbers in peripheral blood in malignant cases. Early detection of the cell in bone marrow is essential as the disease progresses rapidly if left untreated. However, automated classification of the cell is challenging, owing to its fine-grained variability with B-lymphoid precursor cells and imbalanced data points. Deep learning algorithms demonstrate potential for such fine-grained classification as well as suffer from the imbalanced class problem. In this paper, we explore different deep learning-based State-Of-The-Art (SOTA) approaches to tackle imbalanced classification problems. Our experiment includes input, GAN (Generative Adversarial Networks), and loss-based methods to mitigate the issue of imbalanced class on the challenging C-NMC and ALLIDB-2 dataset for leukemia detection. We have shown empirical evidence that loss-based methods outperform GAN-based and input-based methods in imbalanced classification scenarios.

Development of an automated combined positive score prediction pipeline using artificial intelligence on multiplexed immunofluorescence images

Immunotherapy targeting immune checkpoint proteins, such as programmed cell death ligand 1 (PD-L1), has shown impressive outcomes in many clinical trials but only 20%-40% of patients benefit from it. Utilizing Combined Positive Score (CPS) to evaluate PD-L1 expression in tumour biopsies to identify patients with the highest likelihood of responsiveness to anti-PD-1/PD-L1 therapy has been approved by the Food and Drug Administration for several solid tumour types. Current CPS workflow requires a pathologist to manually score the two-colour PD-L1 chromogenic immunohistochemistry image. Multiplex immunofluorescence (mIF) imaging reveals the expression of an increased number of immune markers in tumour biopsies and has been used extensively in immunotherapy research. Recent rapid progress of Artificial Intelligence (AI)-based imaging analysis, particularly Deep Learning, provides cost effective and high-quality solutions to healthcare. In this article, we propose an imaging pipeline that utilizes three-colour mIF images (DAPI, PD-L1, and Pan-cytokeratin) as input and predicts the CPS using AI techniques. Our novel pipeline is composed of three modules employing algorithms of image processing, machine learning, and deep learning techniques. The first module of quality check (QC) detects and removes the image regions contaminated with sectioning and staining artefacts. The QC module ensures that only image regions free of the three common artefacts are used for downstream analysis. The second module of nuclear segmentation uses deep learning to segment and count nuclei in the DAPI images wherein our specialized method can accurately separate touching nuclei. The third module of cell phenotyping calculates CPS by identifying and counting PD-L1 positive cells and tumour cells. These modules are data-efficient and require only few manual annotations for training purposes. Using tumour biopsies from a clinical trial, we found that the CPS from the AI-based models shows a high Spearman correlation (78%, p = 0.003) to the pathologist-scored CPS.

Availability optimization of power generating units used in sewage treatment plants using metaheuristic techniques

Metaheuristic techniques have been utilized extensively to predict industrial systems' optimum availability. This prediction phenomenon is known as the NP-hard problem. Though, most of the existing methods fail to attain the optimal solution due to several limitations like slow rate of convergence, weak computational speed, stuck in local optima, etc. Consequently, in the present study, an effort has been made to develop a novel mathematical model for power generating units assembled in sewage treatment plants. Markov birth-death process is adopted for model development and generation of Chapman-Kolmogorov differential-difference equations. The global solution is discovered using metaheuristic techniques, namely genetic algorithm and particle swarm optimization. All time-dependent random variables associated with failure rates are considered exponentially distributed, while repair rates follow the arbitrary distribution. The repair and switch devices are perfect and random variables are independent. The numerical results of system availability have been derived for different values of crossover, mutation, several generations, damping ratio, and population size to attain optimum value. The results were also shared with plant personnel. Statistical investigation of availability results justifies that particle swarm optimization outdoes genetic algorithm in predicting the availability of power-generating systems. In present study a Markov model is proposed and optimized for performance evaluation of sewage treatment plant. The developed model is one that can be useful for sewage treatment plant designers in establishing new plants and purposing maintenance policies. The same procedure of performance optimization can be adopted in other process industries too.

Learning from dermoscopic images in association with clinical metadata for skin lesion segmentation and classification

Automatic segmentation and classification of lesions are two clinically significant tasks in the computer-aided diagnosis of skin diseases. Both tasks are challenging due to the nonnegligible lesion differences in dermoscopic images from different patients. In this paper, we propose a novel pipeline to efficiently implement skin lesions' segmentation and classification tasks, which consists of a segmentation network and a classification network. To improve the performance of the segmentation network, we propose a novel module of Multi-Scale Holistic Feature Exploration (MSH) to thoroughly exploit perceptual clues latent among multi-scale feature maps as synthesized by the decoder. The MSH module enables holistic exploration of features across multiple scales to more effectively support downstream image analysis tasks. To boost the performance of the classification network, we propose a novel module of Cross-Modality Collaborative Feature Exploration (CMC) to discover latent discriminative features by collaboratively exploiting potential relationships between cross-modal features of dermoscopic images and clinical metadata. The CMC module enables dynamically capturing versatile interaction effects among cross-modal features during the model's representation learning procedure by discriminatively and adaptively learning the interaction weight associated with each crossmodality feature pair. In addition, to effectively reduce background noise and boost the lesion discrimination ability of the classification network, we crop the images based on lesion masks generated by the best segmentation model. We evaluate the proposed pipeline on the four public skin lesion datasets, where the ISIC 2018 and PH2 are for segmentation, and the ISIC 2019 and ISIC 2020 are combined into a new dataset, ISIC 2019&2020, for classification. It achieves a Jaccard index of 83.31% and 90.14% in skin lesion segmentation, an AUC of 97.98% and an Accuracy of 92.63% in skin lesion classification, which is superior to the performance of representative state-of-the-art skin lesion segmentation and classification methods. Last but not least, the new model for segmentation utilizes much fewer model parameters (3.3 M) than its peer approaches, leading to a greatly reduced number of labeled samples required for model training, which obtains substantially stronger robustness than its peers.

Diversity-Based Evolutionary Population Dynamics: A New Operator for Grey Wolf Optimizer

Evolutionary Population Dynamics (EPD) refers to eliminating poor individuals in nature, which is the opposite of survival of the fittest. Although this method can improve the median of the whole population of the meta-heuristic algorithms, it suffers from poor exploration capability to handle high-dimensional problems. This paper proposes a novel EPD operator to improve the search process. In other words, as the primary EPD mainly improves the fitness of the worst individuals in the population, and hence we name it the Fitness-Based EPD (FB-EPD), our proposed EPD mainly improves the diversity of the best individuals, and hence we name it the Diversity-Based EPD (DB-EPD). The proposed method is applied to the Grey Wolf Optimizer (GWO) and named DB-GWO-EPD. In this algorithm, the three most diversified individuals are first identified at each iteration, and then half of the best-fitted individuals are forced to be eliminated and repositioned around these diversified agents with equal probability. This process can free the merged best individuals located in a closed populated region and transfer them to the diversified and, thus, less-densely populated regions in the search space. This approach is frequently employed to make the search agents explore the whole search space. The proposed DB-GWO-EPD is tested on 13 high-dimensional and shifted classical benchmark functions as well as 29 test problems included in the CEC2017 test suite, and four constrained engineering problems. The results obtained by the proposal upon implemented on the classical test problems are compared to GWO, FB-GWO-EPD, and four other popular and newly proposed optimization algorithms, including Aquila Optimizer (AO), Flow Direction Algorithm (FDA), Arithmetic Optimization Algorithm (AOA), and Gradient-based Optimizer (GBO). The experiments demonstrate the significant superiority of the proposed algorithm when applied to a majority of the test functions, recommending the application of the proposed EPD operator to any other meta-heuristic whenever decided to ameliorate their performance.