检索结果-内蒙古大学图书馆

Investigating and modeling day-to-day route choices based on laboratory experiments. Part II: A route-dependent attraction-based stochastic process model

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Qi, Hang Jia, Ning Qu, Xiaobo He, Zhengbing Institute for Advanced Studies in Finance and Economics Hubei University of Economics China Institute of Systems Engineering College of Management and Economics Tianjin University China Laboratory of Computation and Analytics of Complex Management Systems [Tianjin University China State Key Laboratory of Automotive Safety and Energy Tsinghua University China Beijing Key Laboratory of Traffic Engineering Beijing University of Technology China

To explain day-to-day (DTD) route-choice behaviors and traffic dynamics observed in a series of lab experiments, Part I of this research proposed a discrete choice-based analytical dynamic model (Qi et al., 2023). Although the deterministic model could well reproduce the experimental observations, it converges to a stable equilibrium of route flow while the observed DTD evolution is apparently with random oscillations. To overcome the limitation, the paper proposes a route-dependent attraction-based stochastic process (RDAB-SP) model based on the same behavioral assumptions in Part I of this research. Through careful comparison between the model-based estimation and experimental observations, it is demonstrated that the proposed RDAB-SP model can accurately reproduce the random oscillations both in terms of flow switching and route flow evolution. To the best of our knowledge, this is the first attempt to explain and model experimental observations by using stochastic process DTD models, and it is interesting to find that the seemingly unanticipated phenomena (i.e., random route switching behavior) is actually dominated by simple rules, i.e., independent and probability-based route-choice behavior. Finally, an approximated model is developed to help simulate the stochastic process and evaluate the equilibrium distribution in a simple and efficient manner, making the proposed model a useful and practical tool in transportation policy design. © 2023, CC BY.

关键词： Markov processes

Investigating day-to-day route choices based on multi-scenario laboratory experiments. Part I: Route-dependent attraction and its modeling

学校读者我要写书评

暂无评论

arXiv 2023年

In the area of urban transportation networks, a growing number of day-to-day (DTD) traffic dynamic theories have been proposed to describe the network flow evolution, and an increasing amount of laboratory experiments have been conducted to observe travelers’ behavior regularities. However, the "communication" between theorists and experimentalists has not been made well. This paper devotes to 1) detecting unanticipated behavior regularities by conducting a series of laboratory experiments, and 2) improving existing DTD dynamics theories by embedding the observed behavior regularities into a route choice model. First, 312 subjects participated in one of the eight decision-making scenarios and make route choices repeatedly in congestible parallel-route networks. Second, three route-switching behavior patterns that cannot be fully explained by the classic route-choice models are observed. Third, to enrich the explanation power of a discrete route-choice model, behavioral assumptions of route-dependent attractions, i.e., route-dependent inertia and preference, are introduced. An analytical DTD dynamic model is accordingly proposed and proven to steadily converge to a unique equilibrium state. Finally, the proposed DTD model could satisfactorily reproduce the observations in various datasets. The research results can help transportation science theorists to make the best use of laboratory experimentation and to build network equilibrium or DTD dynamic models with both real behavioral basis and neat mathematical properties. © 2023, CC BY.

关键词： Dynamic models

Conflicts and optimization of multiple promotions combinations for merchants

学校读者我要写书评

暂无评论

Expert systems with Applications 2025年 291卷

作者： Rongxiao Huang Zhiyong Zhang Bing Lv Hongke Zhao College of Management and Economics Tianjin University Tianjin 300072 China Laboratory of Computation and Analytics of Complex Management Systems (CACMS) Tianjin 300072 China

Marketing strategies developed by merchants on e-commerce platforms are essential tools to increase sales. To attract consumers, merchants often design multiple promotions for their products and allow them to be stacked. However, combining multiple promotions is not always rational, as some may conflict with each other. In these cases, merchants may suffer unnecessary losses, i.e., the combinations of promotions may have loopholes, which need to be urgently addressed. Thus, this paper discusses the risk assessment and optimal design of promotional combinations. A risk assessment model is established by designing heuristic rules and using machine learning. Using this model, we analyze collected marketing data to identify and categorize various types of promotional loopholes. We then construct corresponding mixed-integer linear programming (MILP) optimization models that aim to maximize merchant profits while minimizing risk. Simulation experiments demonstrate the effectiveness of our predictive model in detecting promotional loopholes. Moreover, our optimization framework provides optimal promotional combination strategies that account for both profitability and risk. Compared to optimizing single promotions in isolation, our approach significantly reduces the likelihood of unreasonable promotional strategies. This provides a practical and effective tool for merchants to assess the risks of their marketing strategies and to optimize promotion parameters before implementation.

关键词：

Use of Machine Learning in Understanding Transport Dynamics of Land Use and Public Transportation in a Developing City

学校读者我要写书评

暂无评论

SSRN

SSRN 2023年

作者： Dorosan, Michael Dailisan, Damian Valenzuela, Jesus Felix Monterola, Christopher Analytics Computing and Complex Systems Laboratory Asian Institute of Management 123 Paseo de Roxas Makati City1229 Philippines Computational Social Science ETH Zürich Stampfenbachstrasse 48 Zürich8092 Switzerland

Rapid urbanization has resulted in problems with pollution, traffic congestion, and public transportation in cities worldwide. Much urbanization research has focused on developing nations, where data is easier to obtain. Our work proposes a machine learning methodology that explores the interplay between land use and public transportation in growing cities, with Cauayan city in the Philippines as the test case. We trained a gradient-boosting tree with GPS tracks, time of day, road properties, and land use data as features to predict vehicle speeds on various roads in a city. Our model results in a root mean squared error of 5.74 km/hr. Apart from predicting speeds, Shapley’s additive explanation (SHAP) values show that proximity to dense commercial and residential areas decreases predicted vehicle speeds. Our approach uses interpretable machine learning models to understand urban transportation in developing cities. © 2023, The Authors. All rights reserved.

关键词： Urban transportation

Correction: Platform pricing and blockchain adoption for capacity sharing with cross-network externality and supply risk

学校读者我要写书评

暂无评论

Annals of Operations Research 2024年第2期341卷 1337-1337页

作者： He, Longfei Zhang, Xiao Huo, Baofeng Zhang, Yang College of Management and Economics & The Laboratory of Computation and Analytics of Complex Management Systems (CACMS) Tianjin University Tianjin China Kent Business School University of Kent Canterbury Kent UK

Deep-learned orthogonal basis patterns for fast, noise-robust single-pixel imaging

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Aguilar, Ritz Ann Dailisan, Damian National Institute of Physics University of the Philippines Diliman Quezon City1101 Philippines Analytics Computing and Complex Systems Laboratory Asian Institute of Management Makati City1229 Philippines

Single-pixel imaging (SPI) is a novel, unconventional method that goes beyond the notion of traditional cameras but can be computationally expensive and slow for real-time applications. Deep learning has been proposed as an alternative approach for solving the SPI reconstruction problem, but a detailed analysis of its performance and generated basis patterns when used for SPI is limited. We present a modified deep convolutional autoencoder network (DCAN) for SPI on 64×64 pixel images with up to 6.25% compression ratio and apply binary and orthogonality regularizers during training. Training a DCAN with these regularizers allows it to learn multiple measurement bases that have combinations of binary or non-binary, and orthogonal or non-orthogonal patterns. We compare the reconstruction quality, orthogonality of the patterns, and robustness to noise of the resulting DCAN models to traditional SPI reconstruction algorithms (such as Total Variation minimization and Fourier Transform). Our DCAN models can be trained to be robust to noise while still having fast enough reconstruction times (∼3 ms per frame) to be viable for real-time imaging. Copyright © 2022, The Authors. All rights reserved.

关键词： Deep learning

Attention Mechanisms Perspective: Exploring LLM Processing of Graph-Structured Data

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Guan, Zhong Wu, Likang Zhao, Hongke He, Ming Fan, Jianping College of Management and Economics Tianjin University Tianjin China Laboratory of Computation and Analytics of Complex Management Systems Tianjin University Tianjin China ai-deepcube China AI Lab at Lenovo Research Beijing China

Attention mechanisms are critical to the success of large language models (LLMs), driving significant advancements in multiple fields. However, for graph-structured data, which requires emphasis on topological connections, they fall short compared to message-passing mechanisms on fixed links, such as those employed by Graph Neural Networks (GNNs). This raises a question: "Does attention fail for graphs in natural language settings?" Motivated by these observations, we embarked on an empirical study from the perspective of attention mechanisms to explore how LLMs process graph-structured data. The goal is to gain deeper insights into the attention behavior of LLMs over graph structures. We uncovered unique phenomena regarding how LLMs apply attention to graph-structured data and analyzed these findings to improve the modeling of such data by LLMs. The primary findings of our research are: 1) While LLMs can recognize graph data and capture text-node interactions, they struggle to model inter-node relationships within graph structures due to inherent architectural constraints. 2) The attention distribution of LLMs across graph nodes does not align with ideal structural patterns, indicating a failure to adapt to graph topology nuances. 3) Neither fully connected attention nor fixed connectivity is optimal;each has specific limitations in its application scenarios. Instead, intermediate-state attention windows improve LLM training performance and seamlessly transition to fully connected windows during inference. Source code: LLM4Exploration © 2025, CC BY.

关键词：

Birthplace bias, familiarity and portfolio choice

学校读者我要写书评

暂无评论

International Review of Financial Analysis 2025年 105卷

作者： Qingchong Chen Xiong Xiong Ya Gao Yumeng Zhang College of Management and Economics Tianjin University Tianjin 300072 China Laboratory of Computation and Analytics of Complex Management Systems (CACMS) Tianjin University Tianjin 300072 China School of Economics and Management Dalian University of Technology Dalian 116024 China

We use high-frequency trading data from individual investors during the period 2012 to 2016 to examine the existence of birthplace bias in China. As we have found, there is a birthplace bias and a local bias in the Chinese stock market and is related to the GDP level of the province, the characteristics of the company, the characteristics of investors, the short-term and long-term state of the market. Meanwhile, when investors move to another province, they maintain their preference for stocks in their birthplace and develop a preference for stocks in their new place of residence. However, these individual investors trade on the basis of familiarity and are unable to achieve excess returns. Thus, the research on proximity bias should not only focus on the existence of proximity bias itself, but also introduce more perspectives for discussion so as to get a more robust conclusion.

关键词： Birthplace bias Familiarity Local bias Portfolio choice

Heterogeneous Graph Reinforcement Learning for Dependency-aware Multi-task Allocation in Spatial Crowdsourcing

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Zhao, Yong Zhu, Zhengqiu Gao, Chen Wang, En Huang, Jincai Wang, Fei-Yue College of Systems Engineering National University of Defense Technology Hunan Province Changsha410073 China BNRist Tsinghua University Beijing100084 China College of Computer Science and Technology Jilin University Changchun130012 China Key Laboratory of Symbolic Computation and Knowledge Engineering of Ministry of Education Jilin University Changchun130012 China State Key Laboratory for Management and Control of Complex Systems Institute of Automation Chinese Academy of Sciences Beijing100190 China

Spatial Crowdsourcing (SC) is gaining traction in both academia and industry, with tasks on SC platforms becoming increasingly complex and requiring collaboration among workers with diverse skills. Recent research works address complex tasks by dividing them into subtasks with dependencies and assigning them to suitable workers. However, the dependencies among subtasks and their heterogeneous skill requirements, as well as the need for efficient utilization of workers’ limited work time in the multi-task allocation mode, pose challenges in achieving an optimal task allocation scheme. Therefore, this paper formally investigates the problem of Dependency-aware Multi-task Allocation (DMA) and presents a well-designed framework to solve it, known as Heterogeneous Graph Reinforcement Learning-based Task Allocation (HGRL-TA). To address the challenges associated with representing and embedding diverse problem instances to ensure robust generalization, we propose a multi-relation graph model and a Compound-path-based Heterogeneous Graph Attention Network (CHANet) for effectively representing and capturing intricate relations among tasks and workers, as well as providing embedding of problem state. The task allocation decision is determined sequentially by a policy network, which undergoes simultaneous training with CHANet using the proximal policy optimization algorithm. Extensive experiment results demonstrate the effectiveness and generality of the proposed HGRL-TA in solving the DMA problem, leading to average profits that is 21.78% higher than those achieved using the metaheuristic methods. © 2024, CC BY-NC-ND.

关键词： Reinforcement learning