Just Adjust One Prompt: Enhancing In-Context Dialogue Scoring Via Constructing The Optimal Subgraph Of Demonstrations And Prompts,
2023
Singapore Management University
Just Adjust One Prompt: Enhancing In-Context Dialogue Scoring Via Constructing The Optimal Subgraph Of Demonstrations And Prompts, Jiashu Pu, Ling Cheng, Lu Fan, Tangjie Lv, Rongsheng Zhang
Research Collection School Of Computing and Information Systems
The use of modern Large Language Models (LLMs) as chatbots still has some problems such as hallucinations and lack of empathy. Identifying these issues can help improve chatbot performance. The community has been continually iterating on reference-free dialogue evaluation methods based on large language models (LLMs) that can be readily applied. However, many of these LLM-based metrics require selecting specific datasets and developing specialized training tasks for different evaluation dimensions (e.g., coherence, informative). The developing step can be time-consuming and may need to be repeated for new evaluation dimensions. To enable efficient and flexible adaptation to diverse needs of dialogue …
Offline Rl With Discrete Proxy Representations For Generalizability In Pomdps,
2023
Singapore Management University
Offline Rl With Discrete Proxy Representations For Generalizability In Pomdps, Pengjie Gu, Xinyu Cai, Dong Xing, Xinrun Wang, Mengchen Zhao, Bo An
Research Collection School Of Computing and Information Systems
Offline Reinforcement Learning (RL) has demonstrated promising results in various applications by learning policies from previously collected datasets, reducing the need for online exploration and interactions. However, real-world scenarios usually involve partial observability, which brings crucial challenges of the deployment of offline RL methods: i) the policy trained on data with full observability is not robust against the masked observations during execution, and ii) the information of which parts of observations are masked is usually unknown during training. In order to address these challenges, we present Offline RL with DiscrEte pRoxy representations (ORDER), a probabilistic framework which leverages novel state …
Controlling Type Confounding In Ad Hoc Teamwork With Instance-Wise Teammate Feedback Rectification,
2023
Singapore Management University
Controlling Type Confounding In Ad Hoc Teamwork With Instance-Wise Teammate Feedback Rectification, Dong Xing, Pengjie Gu, Qian Zheng, Xinrun Wang, Shanqi Liu, Longtao Zheng, Bo An, Gang Pan
Research Collection School Of Computing and Information Systems
Ad hoc teamwork requires an agent to cooperate with unknown teammates without prior coordination. Many works propose to abstract teammate instances into high-level representation of types and then pre-train the best response for each type. However, most of them do not consider the distribution of teammate instances within a type. This could expose the agent to the hidden risk of type confounding. In the worst case, the best response for an abstract teammate type could be the worst response for all specific instances of that type. This work addresses the issue from the lens of causal inference. We first theoretically …
Ethical Considerations For Artificial Intelligence In Educational Assessments,
2023
Singapore Management University
Ethical Considerations For Artificial Intelligence In Educational Assessments, Lim Ming Soon Tristan, Gottipati Swapna, Michelle L. F. Cheong
Research Collection School Of Computing and Information Systems
In the vital context of education, the application of artificial intelligence (AI) to assessments necessitates a nuanced examination of the boundaries between ethically permissible and impermissible practices. In this chapter, the authors applied a systematic literature mapping methodology to scour extant research, so as to holistically structure the landscape into explicit topical research clusters. Through topic modelling and network analyses, research mapped key ethical principles to different assessment phases in a triadic ontological framework. The chapter looks to provide researchers and practitioners the insights into the ethical challenges that exist across an end-to-end assessment pipeline.
(Un)Likelihood Training For Interpretable Embedding,
2023
Singapore Management University
(Un)Likelihood Training For Interpretable Embedding, Jiaxin Wu, Chong-Wah Ngo, Wing-Kwong Chan, Zhijian Hou
Research Collection School Of Computing and Information Systems
Cross-modal representation learning has become a new normal for bridging the semantic gap between text and visual data. Learning modality agnostic representations in a continuous latent space, however, is often treated as a black-box data-driven training process. It is well known that the effectiveness of representation learning depends heavily on the quality and scale of training data. For video representation learning, having a complete set of labels that annotate the full spectrum of video content for training is highly difficult, if not impossible. These issues, black-box training and dataset bias, make representation learning practically challenging to be deployed for video …
Knowing Just Enough To Be Dangerous: The Sociological Effects Of Censoring Public Ai,
2023
Old Dominion University
Knowing Just Enough To Be Dangerous: The Sociological Effects Of Censoring Public Ai, David Hopkins
Cybersecurity Undergraduate Research Showcase
This paper will present the capabilities and security concerns of public AI, also called generative AI, and look at the societal and sociological effects of implementing regulations of this technology.
Application Of Virtual-Real Simulation In Military Field,
2023
College of Intelligence Science and Technology, National University of Defense Technology, Changsha 410000, China
Application Of Virtual-Real Simulation In Military Field, Ziquan Mao, Jialong Gao, Jianxing Gong, Quan Liu
Journal of System Simulation
Abstract: The definition and content of the virtual-real simulation are presented. According to different technical ideas, the development status and existing problems of virtual-real simulation are summarized from three aspects of digital twin, live-virtual-constructive (LVC) simulation, and parallel system. The similarities and differences, as well as the advantages and disadvantages of the three methods are analyzed and compared, and their main application fields are discussed. In order to deal with difficulties encountered in military training, operational tests, equipment development, and equipment maintenance, a solution based on virtual-real simulation is proposed by means of theoretical guidance, case comparison, and transfer and …
Charging Facility Layouts Based On Charging Selection Behavior,
2023
College of Civil Engineering and Architecture, Xinjiang University, Urumqi 830047, China
Charging Facility Layouts Based On Charging Selection Behavior, Lixiao Wang, Zhonghui Wang
Journal of System Simulation
Abstract: A charging facility layout method based on charging selection behavior is proposed for the current uncoordinated development of electric vehicles and charging infrastructure and the low utilization of public charging facilities. The influence of the charging selection behavior of electric vehicle users' trips on the charging facility layout is considered, and a charging selection behavior model is built and applied to the charging demand prediction. Based on the study of charging selection behavior and charging demands, a charging facility layout model with the minimization of total travel time as the objective function is built, and the reciprocal feedback between …
Research On Multi-Process Product Quality Prediction Based On Improved Bilstm,
2023
School of Mechanical Engineering, Shenyang University, Shenyang 110044, China
Research On Multi-Process Product Quality Prediction Based On Improved Bilstm, Tianrui Zhang, Yuting Liu, Yike Wang
Journal of System Simulation
Abstract: In response to the complex manufacturing process of multi-process products, a multi-process product quality prediction model based on the kernel principal component analysis (KPCA) - and improved sparrow search algorithm (ISSA) optimized bi-directional long short term memory (BiLSTM) was proposed to address the uncertain factors that affect product quality, while improving the capacity for each process and ensuring the stability, in multi-process production. Firstly, KPCA was used for data preprocessing, and a kernel function was established on the basis of principal component analysis together with kernel methods. As redundant features were removed through dimension reduction, an improved Gaussian mutation …
Image Semantic Segmentation Algorithm Based On Improved Deeplabv3+,
2023
Liaoning General Aviation Academy, Shenyang Aerospace University, Shenyang 110034, China; College of Electronic Information Engineering, Shenyang Aerospace University, Shenyang 110034, China
Image Semantic Segmentation Algorithm Based On Improved Deeplabv3+, Weiping Zhao, Yu Chen, Song Xiang, Yuanqiang Liu, Chaoyue Wang
Journal of System Simulation
Abstract: Mainstream image semantic segmentation networks currently face problems such as incorrec segmentation, discontinuous segmentation, and high model complexity, which cannot be flexibly and efficiently deployed in practical scenarios. To this end, an image semantic segmentation network that optimizes the DeepLabv3+ model is designed by comprehensively considering the network parameters, prediction time, and accuracy. The lightweight EfficientNetv2 is adopted to extract backbone network features and improve parameter utilization. In the atrous spatial pyramid pooling module, the mixed strip pooling is utilized to replace the global average pooling, and a depthwise separable dilated convolution is introduced to reduce parameters and improve …
Intercell Dynamic Scheduling Method Based On Deep Reinforcement Learning,
2023
University of Shanghai for Science and Technology, Shanghai 200093, China
Intercell Dynamic Scheduling Method Based On Deep Reinforcement Learning, Jing Ni, Mengke Ma
Journal of System Simulation
Abstract: In order to solve the intercell scheduling problem of dynamic arrival of machining tasks and realize adaptive scheduling in the complex and changeable environment of the intelligent factory, a scheduling method based on a deep Q network is proposed. A complex network with cells as nodes and workpiece intercell machining path as directed edges is constructed, and the degree value is introduced to define the state space with intercell scheduling characteristics. A compound scheduling rule composed of a workpiece layer, unit layer, and machine layer is designed, and hierarchical optimization makes the scheduling scheme more global. Since double deep …
Uav-Enabled Task Offloading Strategy For Vehicular Edge Computing Networks,
2023
Nanjing Vocational College of Information Technology, Nanjing 210023, China
Uav-Enabled Task Offloading Strategy For Vehicular Edge Computing Networks, Feng Hu, Haiyang Gu, Jun Lin
Journal of System Simulation
Abstract: As intelligent vehicles are equipped with more and more sensors, the explosive growth of sensor data is generated, which brings severe challenges to vehicular communication and computing. In addition, the modern road presents a three-dimensional structure, and the system architecture of traditional vehicular networks cannot guarantee full coverage and seamless computing. A task offloading strategy for UAV-assisted and 6G-enabled (Sixth Generation) vehicular edge computing networks is proposed. Furthermore, a flexible and intelligent vehicular edge computing mode is composed by vehicles and UAVs, which provide three-dimensional edge computing services for delay-sensitive and computation-intensive vehicular tasks, and ensure timely processing and …
Research On Operational Effectiveness Evaluation Method Of Space-Based Information Support Equipment System,
2023
University of Aerospace Engineering, Beijing 101400, China
Research On Operational Effectiveness Evaluation Method Of Space-Based Information Support Equipment System, Xiaolan Yu, Wei Xiong, Chi Han, Zhenwei Wu
Journal of System Simulation
Abstract: The operational effectiveness evaluation of space-based information support equipment systems has become a research hotspot in the military field. How to effectively deal with the nonlinear and confrontational problems of the operational effectiveness of the space-based information support equipment system has become a crucial issue in the development of the space-based information support equipment system. In this paper, a method for evaluating the operational effectiveness of the space-based information support equipment system based on the system dynamics (SD) model is presented. The SD flow rate basic tree entry modeling method is used to establish the basic tree entry model …
Requirements Of Parallel Combat System Based On Gqfd-Coupling Coordination Degree,
2023
Army Academy of Armored Forces , Beijing 100071, China
Requirements Of Parallel Combat System Based On Gqfd-Coupling Coordination Degree, Zhiming Dong, Bingshan Si, Liang Li
Journal of System Simulation
Abstract: In view of future intelligent unmanned combat characteristics, the concept of parallel combat is proposed and the model of parallel combat system is built based on OODA ring theory. Meanwhile, this paper builds a demand analysis model based on GQFD-coupling coordination degree to solve the low reliability, lack of objectivity, and single description perspective of the traditional quality function deployment (QFD) method and coupling coordination degree analysis during demand analysis. Additionally, the importance ranking of ability requirements in the parallel combat system is obtained by the house of quality of ability requirement analysis in the combat system based on …
Multi-Depot Half-Open Vehicle Routing Problem With Simultaneous Delivery-Pickup And Time Windows,
2023
School of Business Administration, Henan Polytechnic University, Jiaozuo 454003, China
Multi-Depot Half-Open Vehicle Routing Problem With Simultaneous Delivery-Pickup And Time Windows, Yingyu Zhang, Liyun Wu, Shengtai Jia
Journal of System Simulation
Abstract: To solve the multi-depot half-open vehicle routing problem with simultaneous delivery-pickup and time windows, this paper builds a mathematical model of a multi-depot half-open vehicle routing problem with simultaneous delivery-pickup and time windows by balancing the vehicle in and out of the distribution center and minimizing vehicle delivery distance as the goal. According to the characteristics of the problem, a brain storm algorithm based on chaotic mutation is designed to solve this problem,and the sequential crossover strategy is adopted to increase the population diversity. Meanwhile, the algorithm selects two chaotic maps for chaotic mutation operation, which employs the diversity, …
Learning-Based Ant Colony Optimization Algorithm For Solving A Kind Of Complex
2-Echelon Vehicle Routing Problem,
2023
School of Information Engineering and Automation, Kunming University of Science and Technology, Kunming 650500, China
Learning-Based Ant Colony Optimization Algorithm For Solving A Kind Of Complex 2-Echelon Vehicle Routing Problem, Xue Chen, Rong Hu, Hui Wang, Zuocheng Li, Bin Qian, Yixu Li
Journal of System Simulation
Abstract: Aiming at green 2-echelon vehicle routing problem with simultaneous pick-up and delivery, a learning-based ant colony optimization algorithm combined with clustering decomposition is proposed. The objective function to be minimized is total transportation cost wherein carbon emission cost is specially considered. Associated with the mutual coupling features of the 2-echelon vehicle routing problem, we propose a distance-based clustering method to decompose the original problem into a set of sub-problems. Then, a learning-based ant colony optimization algorithm is presented to find the solutions of the sub-problems based on which the solution of the original problem can be obtained. In the …
Analysis Of Autonomous Aerial Refueling Capability Requirements And Key Evaluation
Indicators,
2023
Chinese Flight Test Establishment, Xi'an 710089, China; Key Laboratory of Flight Simulation Aeronautical Technology of AVIC, Xi'an 710089, China
Analysis Of Autonomous Aerial Refueling Capability Requirements And Key Evaluation Indicators, Quan Zou, Yixin Hua, Zhu Shao, Wenbi Zhao
Journal of System Simulation
Abstract: From the perspective of flight tests, how to evaluate the autonomous aerial refueling (AAR) capability and select key indicators for evaluation is a key problem to be solved for AAR trials. The standards requirements of aerial refueling and manned aircraft aerial refueling experience in China and abroad are analyzed. The total capability of AAR is studied, and key evaluation indicators in the AAR whole process including rendezvous, formation, docking, refueling, and disengagement are proposed. The evaluation method is demonstrated in both numerical simulation and hardware-in-loop test environments. Finally, the key indicators affecting the docking success of AAR are analyzed, …
Imitative Generation Of Optimal Guidance Law Based On Reinforcement Learning,
2023
Beijing Simulation Center, Beijing 100854, China
Imitative Generation Of Optimal Guidance Law Based On Reinforcement Learning, Zhengxuan Jia, Tingyu Lin, Yingying Xiao, Guoqiang Shi, Hao Wang, Bi Zeng, Yiming Ou, Pengpeng Zhao
Journal of System Simulation
Abstract: Under the background of high-speed maneuvering target interception, an optimal guidance law generation method for head-on interception independent of target acceleration estimation is proposed based on deep reinforcement learning. In addition, its effectiveness is verified through simulation experiments. As the simulation results suggest, the proposed method successfully achieves head-on interception of high-speed maneuvering targets in 3D space and largely reduces the requirement for target estimation with strong uncertainty, and it is more applicable than the optimal control method.
Development Of Combat Concept Of Intelligent Land Assault System Based On Dodaf,
2023
Army Academy of Armored Forces, Beijing 100072, China
Development Of Combat Concept Of Intelligent Land Assault System Based On Dodaf, Can Wang, Haoran Ji, Qisheng Guo, Zhiming Dong, Yaxin Tan, Ge Mu
Journal of System Simulation
Abstract: In view of military demand traction in the development of land assault equipment, a combat concept of land assault systems for future intelligent combat is developed. Basedon the definition of relevant concepts and research boundaries, the combat concept model framework and modeling steps are proposed based on DoDAF, and the combat effect, combat process, combat nodes, resource interaction, system composition, and capability characteristics are analyzed in combination with the model description. The combat concept verification is carried out from the aspects of system combat efficiency and communication load by simulation experiments. The results show that the intelligent assault system …
Adaptive Robust Trajectory Tracking Control For Nsv With Multiple Stochastic Disturbances,
2023
Key Laboratory of Applied Mathematics and Artificial Intelligence Mechanism, Hefei University, Hefei 230601, China; Hefei Comprehensive National Science Center, Institute of Artificial Intelligence, Hefei 230088, China
Adaptive Robust Trajectory Tracking Control For Nsv With Multiple Stochastic Disturbances, Xiaohu Yan, Yuwu Yao, Yuhua Wu, Jiangxin Xu
Journal of System Simulation
Abstract: A stochastic control scheme of adaptive robust trajectory tracking is proposed for near space vehicle (NSV) with stochastic noise input disturbances, Poisson random fluctuation disturbances, and control input saturation. The effective tracking of the height and speed reference signals is realized. For the outer loop trajectory control, the robust stochastic controller is designed for the height subsystem and the speed subsystem respectively. Additionally, the required attitude angle reference signals for the inner loop attitude control are obtained by converting the equivalent control input via numerical calculation. For the inner loop attitude control problems, an adaptive robust stochastic control scheme …
