A Multi-Task Deep Learning Framework for Simultaneous Detection of Thoracic Pathology through Image Classification

Thoracic diseases pose significant risks to an individual's chest health and are among the most perilous medical diseases. They can impact either one or both lungs, which leads to a severe impairment of a person’s ability to breathe normally. Some notable examples of such diseases encompass pneumonia, lung cancer, coronavirus disease 2019 (COVID-19), tuberculosis, and chronic obstructive pulmonary disease (COPD). Consequently, early and precise detection of these diseases is paramount during the diagnostic process. Traditionally, the primary methods employed for the detection involve the use of X-ray imaging or computed tomography (CT) scans. Nevertheless, due to the scarcity of proficient radiologists and the inherent similarities between these diseases, the accuracy of detection can be compromised, leading to imprecise or erroneous results. To address this challenge, scientists have turned to computer-based solutions, aiming for swift and accurate diagnoses. The primary objective of this study is to develop two machine learning models, utilizing single-task and multi-task learning frameworks, to enhance classification accuracy. Within the multi-task learning architecture, two principal approaches exist soft parameter sharing and hard parameter sharing. Consequently, this research adopts a multi-task deep learning approach that leverages CNNs to achieve improved classification performance for the specified tasks. These tasks, focusing on pneumonia and COVID-19, are processed and learned simultaneously within a multi-task model. To assess the effectiveness of the trained model, it is rigorously validated using three different real-world datasets for training and testing.

Policy Network-Based Dual-Agent Deep Reinforcement Learning for Multi-Resource Task Offloading in Multi-Access Edge Cloud Networks

The Multi-access Edge Cloud(MEC) networks extend cloud computing services and capabilities to the edge of the networks. By bringing computation and storage capabilities closer to end-users and connected devices, MEC networks can support a wide range of applications. MEC networks can also leverage various types of resources, including computation resources, network resources, radio resources,and location-based resources, to provide multidimensional resources for intelligent applications in 5/6G.However, tasks generated by users often consist of multiple subtasks that require different types of resources. It is a challenging problem to offload multiresource task requests to the edge cloud aiming at maximizing benefits due to the heterogeneity of resources provided by devices. To address this issue,we mathematically model the task requests with multiple subtasks. Then, the problem of task offloading of multi-resource task requests is proved to be NP-hard. Furthermore, we propose a novel Dual-Agent Deep Reinforcement Learning algorithm with Node First and Link features(NF_L_DA_DRL) based on the policy network, to optimize the benefits generated by offloading multi-resource task requests in MEC networks. Finally, simulation results show that the proposed algorithm can effectively improve the benefit of task offloading with higher resource utilization compared with baseline algorithms.

Task assignment in ground-to-air confrontation based on multiagent deep reinforcement learning

The scale of ground-to-air confrontation task assignments is large and needs to deal with many concurrent task assignments and random events.Aiming at the problems where existing task assignment methods are applied to ground-to-air confrontation,there is low efficiency in dealing with complex tasks,and there are interactive conflicts in multiagent systems.This study proposes a multiagent architecture based on a one-general agent with multiple narrow agents(OGMN)to reduce task assignment conflicts.Considering the slow speed of traditional dynamic task assignment algorithms,this paper proposes the proximal policy optimization for task assignment of general and narrow agents(PPOTAGNA)algorithm.The algorithm based on the idea of the optimal assignment strategy algorithm and combined with the training framework of deep reinforcement learning(DRL)adds a multihead attention mechanism and a stage reward mechanism to the bilateral band clipping PPO algorithm to solve the problem of low training efficiency.Finally,simulation experiments are carried out in the digital battlefield.The multiagent architecture based on OGMN combined with the PPO-TAGNA algorithm can obtain higher rewards faster and has a higher win ratio.By analyzing agent behavior,the efficiency,superiority and rationality of resource utilization of this method are verified.

基于Multi-Agent的无人机集群体系自主作战系统设计

针对无人集群自主作战体系设计中的关键问题,提出基于Multi-Agent的无人集群自主作战系统设计方法。建立无人集群各节点的Agent模型及其推演规则;对于仿真系统模块化和通用化的需求,设计系统互操作式接口和无人集群自主作战的交互关系;开展无人集群系统仿真推演验证。仿真结果表明,所提设计方案不仅能够有效开展并完成自主作战网络生成-集群演化-效能评估的全过程动态演示验证,而且能够通过重复随机试验进一步评估无人集群的协同作战效能,最后总结了集群协同作战的策略和经验。

Evolutionary Multitasking With Global and Local Auxiliary Tasks for Constrained Multi-Objective Optimization

Constrained multi-objective optimization problems(CMOPs) include the optimization of objective functions and the satisfaction of constraint conditions, which challenge the solvers.To solve CMOPs, constrained multi-objective evolutionary algorithms(CMOEAs) have been developed. However, most of them tend to converge into local areas due to the loss of diversity. Evolutionary multitasking(EMT) is new model of solving complex optimization problems, through the knowledge transfer between the source task and other related tasks. Inspired by EMT, this paper develops a new EMT-based CMOEA to solve CMOPs, in which the main task, a global auxiliary task, and a local auxiliary task are created and optimized by one specific population respectively. The main task focuses on finding the feasible Pareto front(PF), and global and local auxiliary tasks are used to respectively enhance global and local diversity. Moreover, the global auxiliary task is used to implement the global search by ignoring constraints, so as to help the population of the main task pass through infeasible obstacles. The local auxiliary task is used to provide local diversity around the population of the main task, so as to exploit promising regions. Through the knowledge transfer among the three tasks, the search ability of the population of the main task will be significantly improved. Compared with other state-of-the-art CMOEAs, the experimental results on three benchmark test suites demonstrate the superior or competitive performance of the proposed CMOEA.

Complex task planning method of space-aeronautics cooperative observation based on multi-layer interaction

With the new development trend of multi-resource coordinated Earth observation and the new goal of Earth observation application of“short response time,high observation accuracy,and wide coverage”,space-aeronautics cooperative complex task planning problem has become an urgent problem to be solved.The focus of this problem is to use multiple resources to perform collaborative observations on complex tasks.By analyzing the process from task assignment to receiving task observation results,we propose a multi-layer interactive task planning framework which is composed of a preprocessing method for complex tasks,a task allocation layer,a task planning layer,and a task coordination layer.According to the characteristics of the framework,a hybrid genetic parallel tabu(HGPT)algorithm is proposed on this basis.The algorithm uses genetic annealing algorithm(GAA),parallel tabu(PT)algorithm,and heuristic rules to achieve task allocation,task planning,and task coordination.At the same time,coding improvements,operator design,annealing operations,and parallel calculations are added to the algorithm.In order to verify the effectiveness of the algorithm,simulation experiments under complex task scenarios of different scales are carried out.Experimental results show that this method can effectively solve the problems of observing complex tasks.Meanwhile,the optimization effect and convergence speed of the HGPT is better than that of the related algorithms.

Survey on Task Scheduling Optimization Strategy under Multi-Cloud Environment

Cloud computing technology is favored by users because of its strong computing power and convenient services.At the same time,scheduling performance has an extremely efficient impact on promoting carbon neutrality.Currently,scheduling research in the multi-cloud environment aims to address the challenges brought by business demands to cloud data centers during peak hours.Therefore,the scheduling problem has promising application prospects under themulti-cloud environment.This paper points out that the currently studied scheduling problems in the multi-cloud environment mainly include independent task scheduling and workflow task scheduling based on the dependencies between tasks.This paper reviews the concepts,types,objectives,advantages,challenges,and research status of task scheduling in the multi-cloud environment.Task scheduling strategies proposed in the existing related references are analyzed,discussed,and summarized,including research motivation,optimization algorithm,and related objectives.Finally,the research status of the two kinds of task scheduling is compared,and several future important research directions of multi-cloud task scheduling are proposed.

A Trusted Multi-Task Distribution Mechanism for Internet of Vehicles Based on Smart Contract

In this paper,a trusted multi-task distribution mechanism for Internet of Vehicles based on smart contract is proposed to improve the security and efficiency for the task distribution in Internet of Vehicles.Firstly,a three-tier trusted multi-task distribution framework is presented based on smart contract.The smart contract will be triggered by the task request.As the important part of the smart contract,the task distribution algorithm is stored on the blockchain and run automatically.In the process of the task distribution,the cost of the task distribution and the system stability play a critical role.Therefore,the task distribution problem is formulated to minimize the cost of the task distribution whilst maintaining the stability of the system based on Lyapunov theorem.Unfortunately,this problem is a mixed integer nonlinear programming problem with NP-hard characteristics.To tackle this,the optimization problem is decomposed into two sub problems of computing resource allocation and task distribution decision,and an effective task distribution algorithm is proposed.Simulation results show that the proposed algorithm can effectively improves system performance.

A New Partial Task Offloading Method in a Cooperation Mode under Multi-Constraints for Multi-UE

In Multi-access Edge Computing(MEC),to deal with multiple user equipment(UE)’s task offloading problem of parallel relationships under the multi-constraints,this paper proposes a cooperation partial task offloading method(named CPMM),aiming to reduce UE’s energy and computation consumption,while meeting the task completion delay as much as possible.CPMM first studies the task offloading of single-UE and then considers the task offloading ofmulti-UE based on single-UE task offloading.CPMMuses the critical path algorithmto divide the modules into key and non-key modules.According to some constraints of UE-self when offloading tasks,it gives priority to non-key modules for offloading and uses the evaluation decision method to select some appropriate key modules for offloading.Based on fully considering the competition between multiple UEs for communication resources and MEC service resources,CPMM uses the weighted queuing method to alleviate the competition for communication resources and uses the branch decision algorithm to determine the location of module offloading by BS according to the MEC servers’resources.It achieves its goal by selecting reasonable modules to offload and using the cooperation ofUE,MEC,andCloudCenter to determine the execution location of themodules.Extensive experiments demonstrate that CPMM obtains superior performances in task computation consumption reducing around 6%on average,task completion delay reducing around 5%on average,and better task execution success rate than other similar methods.

Multi-Task Timing Assignment Algorithm for Intelligent Production of Vegetables in Open Field

Vegetable production in the open field involves many tasks,such as soil preparation,ridging,and transplanting/sowing.Different tasks require agricultural machinery equipped with different agricultural tools to meet the needs of the operation.Aiming at the coupling multi-task in the intelligent production of vegetables in the open field,the task assignment method for multiple unmanned tractors based on consistency alliance is studied.Firstly,unmanned vegetable production in the open field is abstracted as a multi-task assignment model with constraints of task demand,task sequence,and the distance traveled by an unmanned tractor.The tight time constraints between associated tasks are transformed into time windows.Based on the driving distance of the unmanned tractor and the replacement cost of the tools,an expanded task cost function is innovatively established.The task assignment model of multiple unmanned tractors is optimized by the consensus based bundle algorithm(CBBA)with time windows.Experiments show that the method can effectively solve task conflict in unmanned production and optimize task allocation.A basic model is provided for the cooperative task of multiple unmanned tractors for vegetable production in the open field.

Satellite-Air-Terrestrial Cloud Edge Collaborative Networks:Architecture,Multi-Node Task Processing and Computation

Integrated satellite-terrestrial network(ISTN)has been considered a novel network architecture to achieve global three-dimensional coverage and ultra-wide area broadband access anytime and anywhere.Being a promising paradigm,cloud computing and mobile edge computing(MEC)have been identified as key technology enablers for ISTN to further improve quality of service and business continuity.However,most of the existing ISTN studies based on cloud computing and MEC regard satellite networks as relay networks,ignoring the feasibility of directly deploying cloud computing nodes and edge computing nodes on satellites.In addition,most computing tasks are transferred to cloud servers or offloaded to nearby edge servers,the layered design of integrated satellite-air-terrestrial architecture and the cloud-edge-device cooperative processing problems have not been fully considered.Therefore,different from previous works,this paper proposed a novel satellite-air-terrestrial layered architecture for cloud-edge-device collaboration,named SATCECN.Then this paper analyzes the appropriate deployment locations of cloud servers and edge servers in ISTN,and describes the processing flow of typical satellite computing tasks.For computing resource allocation problems,this paper proposed a device-edge-cloud Multi-node Cross-layer Collaboration Computing(MCCC)method to find the optimal task allo-cation strategy that minimizes the task completion delay and the weighted system energy consumption.Furthermore,the approximate optimal solutions of the optimization model are obtained by using successive convex approxi-mation algorithm,and the outstanding advantages of the proposed method in reducing system energy consumption and task execution delay are verified through experiments.Finally,some potential issues and directions for future research are highlighted.

PLDMLT:Multi-Task Learning of Diabetic Retinopathy Using the Pixel-Level Labeled Fundus Images

In the field of medical images,pixel-level labels are time-consuming and expensive to acquire,while image-level labels are relatively easier to obtain.Therefore,it makes sense to learn more information(knowledge)from a small number of hard-to-get pixel-level annotated images to apply to different tasks to maximize their usefulness and save time and training costs.In this paper,using Pixel-Level Labeled Images forMulti-Task Learning(PLDMLT),we focus on grading the severity of fundus images for Diabetic Retinopathy(DR).This is because,for the segmentation task,there is a finely labeled mask,while the severity grading task is without classification labels.To this end,we propose a two-stage multi-label learning weakly supervised algorithm,which generates initial classification pseudo labels in the first stage and visualizes heat maps at all levels of severity using Grad-Cam to further provide medical interpretability for the classification task.A multitask model framework with U-net as the baseline is proposed in the second stage.A label update network is designed to alleviate the gradient balance between the classification and segmentation tasks.Extensive experimental results show that our PLDMLTmethod significantly outperforms other stateof-the-art methods in DR segmentation on two public datasets,achieving up to 98.897%segmentation accuracy.In addition,our method achieves comparable competitiveness with single-task fully supervised learning in the DR severity grading task.

Improved Scatter Search Algorithm for Multi-skilled Personnel Scheduling of Ship Block Painting

This paper focuses on the optimization method for multi-skilled painting personnel scheduling.The budget working time analysis is carried out considering the influence of operating area,difficulty of spraying area,multi-skilled workers,and worker’s efficiency,then a mathematical model is established to minimize the completion time. The constraints of task priority,paint preparation,pump management,and neighbor avoidance in the ship block painting production are considered. Based on this model,an improved scatter search(ISS)algorithm is designed,and the hybrid approximate dynamic programming(ADP)algorithm is used to improve search efficiency. In addition,the two solution combination methods of path-relinking and task sequence combination are used to enhance the search breadth and depth. The numerical experimental results show that ISS has a significant advantage in solving efficiency compared with the solver in small scale instances;Compared with the scatter search algorithm and genetic algorithm,ISS can stably improve the solution quality. Verified by the production example,ISS effectively shortens the total completion time of the production,which is suitable for scheduling problems in the actual painting production of the shipyard.

Joint Task Allocation and Resource Optimization for Blockchain Enabled Collaborative Edge Computing

Collaborative edge computing is a promising direction to handle the computation intensive tasks in B5G wireless networks.However,edge computing servers(ECSs)from different operators may not trust each other,and thus the incentives for collaboration cannot be guaranteed.In this paper,we propose a consortium blockchain enabled collaborative edge computing framework,where users can offload computing tasks to ECSs from different operators.To minimize the total delay of users,we formulate a joint task offloading and resource optimization problem,under the constraint of the computing capability of each ECS.We apply the Tammer decomposition method and heuristic optimization algorithms to obtain the optimal solution.Finally,we propose a reputation based node selection approach to facilitate the consensus process,and also consider a completion time based primary node selection to avoid monopolization of certain edge node and enhance the security of the blockchain.Simulation results validate the effectiveness of the proposed algorithm,and the total delay can be reduced by up to 40%compared with the non-cooperative case.

Task Offloading in Edge Computing Using GNNs and DQN

In a network environment composed of different types of computing centers that can be divided into different layers(clod,edge layer,and others),the interconnection between them offers the possibility of peer-to-peer task offloading.For many resource-constrained devices,the computation of many types of tasks is not feasible because they cannot support such computations as they do not have enough available memory and processing capacity.In this scenario,it is worth considering transferring these tasks to resource-rich platforms,such as Edge Data Centers or remote cloud servers.For different reasons,it is more exciting and appropriate to download various tasks to specific download destinations depending on the properties and state of the environment and the nature of the functions.At the same time,establishing an optimal offloading policy,which ensures that all tasks are executed within the required latency and avoids excessive workload on specific computing centers is not easy.This study presents two alternatives to solve the offloading decision paradigm by introducing two well-known algorithms,Graph Neural Networks(GNN)and Deep Q-Network(DQN).It applies the alternatives on a well-known Edge Computing simulator called PureEdgeSimand compares them with the two defaultmethods,Trade-Off and Round Robin.Experiments showed that variants offer a slight improvement in task success rate and workload distribution.In terms of energy efficiency,they provided similar results.Finally,the success rates of different computing centers are tested,and the lack of capacity of remote cloud servers to respond to applications in real-time is demonstrated.These novel ways of finding a download strategy in a local networking environment are unique as they emulate the state and structure of the environment innovatively,considering the quality of its connections and constant updates.The download score defined in this research is a crucial feature for determining the quality of a download path in the GNN training process and has not previously been proposed.Simultaneously,the suitability of Reinforcement Learning(RL)techniques is demonstrated due to the dynamism of the network environment,considering all the key factors that affect the decision to offload a given task,including the actual state of all devices.

Multi-Agent模式下的城市暴雨内涝应急决策方法研究

为厘清应对暴雨内涝灾害动态决策过程中决策主体、决策、决策方案等决策要素间的不确定关系,提出1种多主体(Multi-Agent)和贝叶斯决策网络(BDN)相结合的应急决策方法。首先分阶段构建“主体-任务”可视化网络,分析暴雨内涝灾害各应急阶段的主要任务和参与的决策主体;在考虑到决策要素间的动态不确定性可能造成决策风险的前提下,运用Multi-Agent和BDN方法探究各决策要素间的影响关系,以便进行方案集优选。研究结果表明:该方法具有实用性和现实意义,研究结果可为城市暴雨内涝灾害的应急决策提供理论参考。

Multi-scale data joint inversion of minerals and porosity in altered igneous reservoirs—A case study in the South China Sea

There are abundant igneous gas reservoirs in the South China Sea with significant value of research,and lithology classification,mineral analysis and porosity inversion are important links in reservoir evaluation.However,affected by the diverse lithology,complicated mineral and widespread alteration,conventional logging lithology classification and mineral inversion become considerably difficult.At the same time,owing to the limitation of the wireline log response equation,the quantity and accuracy of minerals can hardly meet the exploration requirements of igneous formations.To overcome those issues,this study takes the South China Sea as an example,and combines multi-scale data such as micro rock slices,petrophysical experiments,wireline log and element cutting log to establish a set of joint inversion methods for minerals and porosity of altered igneous rocks.Specifically,we define the lithology and mineral characteristics through core slices and mineral data,and establish an igneous multi-mineral volumetric model.Then we determine element cutting log correction method based on core element data,and combine wireline log and corrected element cutting log to perform the lithology classification and joint inversion of minerals and porosity.However,it is always difficult to determine the elemental eigenvalues of different minerals in inversion.This paper uses multiple linear regression methods to solve this problem.Finally,an integrated inversion technique for altered igneous formations was developed.The results show that the corrected element cutting log are in good agreement with the core element data,and the mineral and porosity results obtained from the joint inversion based on the wireline log and corrected element cutting log are also in good agreement with the core data from X-ray diffraction.The results demonstrate that the inversion technique is applicable and this study provides a new direction for the mineral inversion research of altered igneous formations.

Associative Tasks Computing Offloading Scheme in Internet of Medical Things with Deep Reinforcement Learning

The Internet of Medical Things(Io MT) is regarded as a critical technology for intelligent healthcare in the foreseeable 6G era. Nevertheless, due to the limited computing power capability of edge devices and task-related coupling relationships, Io MT faces unprecedented challenges. Considering the associative connections among tasks, this paper proposes a computing offloading policy for multiple-user devices(UDs) considering device-to-device(D2D) communication and a multi-access edge computing(MEC)technique under the scenario of Io MT. Specifically,to minimize the total delay and energy consumption concerning the requirement of Io MT, we first analyze and model the detailed local execution, MEC execution, D2D execution, and associated tasks offloading exchange model. Consequently, the associated tasks’ offloading scheme of multi-UDs is formulated as a mixed-integer nonconvex optimization problem. Considering the advantages of deep reinforcement learning(DRL) in processing tasks related to coupling relationships, a Double DQN based associative tasks computing offloading(DDATO) algorithm is then proposed to obtain the optimal solution, which can make the best offloading decision under the condition that tasks of UDs are associative. Furthermore, to reduce the complexity of the DDATO algorithm, the cacheaided procedure is intentionally introduced before the data training process. This avoids redundant offloading and computing procedures concerning tasks that previously have already been cached by other UDs. In addition, we use a dynamic ε-greedy strategy in the action selection section of the algorithm, thus preventing the algorithm from falling into a locally optimal solution. Simulation results demonstrate that compared with other existing methods for associative task models concerning different structures in the Io MT network, the proposed algorithm can lower the total cost more effectively and efficiently while also providing a tradeoff between delay and energy consumption tolerance.

Integration of Multiple Spectral Data via a Logistic Regression Algorithm for Detection of Crop Residue Burned Areas:A Case Study of Songnen Plain,Northeast China

The burning of crop residues in fields is a significant global biomass burning activity which is a key element of the terrestrial carbon cycle,and an important source of atmospheric trace gasses and aerosols.Accurate estimation of cropland burned area is both crucial and challenging,especially for the small and fragmented burned scars in China.Here we developed an automated burned area mapping algorithm that was implemented using Sentinel-2 Multi Spectral Instrument(MSI)data and its effectiveness was tested taking Songnen Plain,Northeast China as a case using satellite image of 2020.We employed a logistic regression method for integrating multiple spectral data into a synthetic indicator,and compared the results with manually interpreted burned area reference maps and the Moderate-Resolution Imaging Spectroradiometer(MODIS)MCD64A1 burned area product.The overall accuracy of the single variable logistic regression was 77.38%to 86.90%and 73.47%to 97.14%for the 52TCQ and 51TYM cases,respectively.In comparison,the accuracy of the burned area map was improved to 87.14%and 98.33%for the 52TCQ and 51TYM cases,respectively by multiple variable logistic regression of Sentind-2 images.The balance of omission error and commission error was also improved.The integration of multiple spectral data combined with a logistic regression method proves to be effective for burned area detection,offering a highly automated process with an automatic threshold determination mechanism.This method exhibits excellent extensibility and flexibility taking the image tile as the operating unit.It is suitable for burned area detection at a regional scale and can also be implemented with other satellite data.

Incidence, risk factors and clinical outcome of multidrug-resistant organisms after heart transplantation

BACKGROUND Transplant recipients commonly harbor multidrug-resistant organisms(MDROs),as a result of frequent hospital admissions and increased exposure to antimi-crobials and invasive procedures.AIM To investigate the impact of patient demographic and clinical characteristics on MDRO acquisition,as well as the impact of MDRO acquisition on intensive care unit(ICU)and hospital length of stay,and on ICU mortality and 1-year mortality post heart transplantation.METHODS This retrospective cohort study analyzed 98 consecutive heart transplant patients over a ten-year period(2013-2022)in a single transplantation center.Data was collected regarding MDROs commonly encountered in critical care.RESULTS Among the 98 transplanted patients(70%male),about a third(32%)acquired or already harbored MDROs upon transplantation(MDRO group),while two thirds did not(MDRO-free group).The prevalent MDROs were Acinetobacter baumannii(14%),Pseudomonas aeruginosa(12%)and Klebsiella pneumoniae(11%).Compared to MDRO-free patients,the MDRO group was characterized by higher body mass index(P=0.002),higher rates of renal failure(P=0.017),primary graft dysfunction(10%vs 4.5%,P=0.001),surgical re-exploration(34%vs 14%,P=0.017),mechanical circulatory support(47%vs 26%P=0.037)and renal replacement therapy(28%vs 9%,P=0.014),as well as longer extracorporeal circulation time(median 210 vs 161 min,P=0.003).The median length of stay was longer in the MDRO group,namely ICU stay was 16 vs 9 d in the MDRO-free group(P=0.001),and hospital stay was 38 vs 28 d(P=0.006),while 1-year mortality was higher(28%vs 7.6%,log-rank-χ2:7.34).CONCLUSION Following heart transplantation,a predominance of Gram-negative MDROs was noted.MDRO acquisition was associated with higher complication rates,prolonged ICU and total hospital stay,and higher post-transplantation mortality.