Embedded Files
Ongoing Collaboration
👉 Ongoing Funded Projects
PI, "Data-Driven Inventory Management with Financial Hedging," General Research Fund (GRF), 2025.1.1- 2026.12.31
PI, "Data-Driven Inventory, Pricing, and Promotion Optimization with Proxy Covariates/ 基于数据驱动和代理特征的库存、定价与促销优化研究," Science Fund for Young Scholars, NSFC, 2025.1.1-2027.12.31
PI, Start-Up Fund for Research, 2025.8.1-2028.7.31
PI, Seed Fund for Basic Research, 2025.8.1 - 2027.7.31
PI, Research Development Fund at DASE Dept. at HKU
Co-PI (¥ 1,440,000), New-generation Smart Manufacturing ERP, ¥ 27,000,000, 国家重点研发计划 重点专项, 2024.12 - 2027.11
Deputy PC/Co-PI, "Planning and Scheduling of Automated Material Handling Systems in Semiconductor Manufacturing Environments," 2025.10-2030.9. HK$ 18,000,000+.
Deputy PC/Co-I, "Coordinated Emergency Evacuation of Multimodal Transportation Systems via Intelligent Navigation Applications 突發事件下基於智慧導航的多模式交通協同疏散研究," 国家重点研发计划项目 Mainland-Hong Kong Joint Funding Scheme (MHKJFS), Innovation and Technology Fund+ Sponsorship from Amap, 2024.6.1 - 2026.5.31 (PI: Prof. Max Shen, Mainland Partner: Tsinghua University)
Co-I, "Healthcare resources allocation: targeting the right intervention to the right population," General Research Fund (GRF), 2025.1.1- 2027.12.31 (PI: Prof. Frank Chen)
Co-I, 国家自然科学基金委员会 (NSFC), 面上项目, "分级诊疗制度下考虑重点患者就医需求的重大传染病应急医疗资源配置," 2024.1.1-2027.12.31
👉 S. Lin, Y. F. Chen, Y. Li, and Z.-J. M. Shen, “Data-driven newsvendor problems regularized by a profit risk constraint,” Production and Operations Management, vol. 31, pp. 1630-1644, Apr 2022. Available at SSRN.
Data Source: AsiaPlayer, a CD/DVD retailer in HK
本文為新品庫存管理問題提出了一個基於機器學習的數據驅動決策分析框架，能夠根據相似產品的歷史信息為新品決定最佳訂貨量。本文是第一篇為風險規避模型和機器學習正則化建立聯繫的文章，數據驅動模型引入了風險約束來正則化訂貨量決策，同時保證了數據驅動求解方案的漸進最優性能。文章發現了一個有趣的啟示：與理論上的風險規避模型恰恰相反，在數據驅動決策下，平均利潤可能受益於較強的風險規避程度。此外，即使是風險中性的決策者也可能從價值風險約束中受益。
👉 S. Lin, M. Bakker, and E. Leung, “Robust meal delivery service for the elderly: a case study in Hong Kong,” Nature Research Journal - Scientific Reports, vol. 12(1), pp. 1-12, Dec 2022. Link.
Data Source: Integrated Home Care Services Teams in HK
本文專門針對香港的社區送餐服務提出了一種本土化的兩階車輛路線模型（送餐人員先從社區中心乘車到各個中轉站，再步行到達長者家中），同時考慮了交付效率和員工工作量的公平性，並刻畫了很多現實特徵，包括護理連續性、香港高密的建築分佈特徵、被服務老年人的特性等。本文採用魯棒優化模型來區分體弱老年人和普通老年人的服務時間不確定性，並提供兩個容易求解的分解算法。通過真實數據案例研究表明，魯棒優化解決方案可以實現更好的性能，產生更少的額外成本，本文還為從業者決策和人力資源規劃提供運營啟示。
👉 S. Lin, Q. Zhang, Y. F. Chen, L. Luo, L. Chen, and W. Zhang, “Smooth Bayesian network model for the prediction of future high-cost patients with COPD,” International Journal of Medical Informatics, vol. 126, pp. 147-155, Apr 2019. Link.
Data Source: Medical Insurance Data in a large city
本文提出了一種混合機器學習預測模型來識別未來高花費的慢性阻塞性肺病患者。構建的貝葉斯模型考慮了帶關因果關係的領域知識，透過網路流優化模型，能生成不同高花費風險因子的因果圖結構，預測準確率比其它機器學習標準方法高。
Page updated
Google Sites
Report abuse