Fashion apparel companies require planning for the next season, a year in advance for supply chain management. This study focuses on size selection decision making for Levi Strauss. Currently, the region and planning group level size grids are built and managed manually. The company suffers from the workload it creates for sizing, merchant and planning teams. This research is aiming to answer two research questions: "Which sizes should be available to the planners under each size grid name for the next season(s)?" and "Which sizes should be adopted for each planning group for the next season(s)?". We approach to the problem with a classification model, which is one of the popular models used in machine learning. With this research, a more automated process was created by using machine learning techniques. A decrease in workload of the teams in the company is expected after it is put into practice. Unlike many studies in the state of art for fashion and apparel industry, this study focuses on sizes where the stock keeping unit represents a product with a certain size.

翻译：服装企业需提前一年为下一季进行供应链规划。本研究聚焦于Levi Strauss的尺码选择决策问题。目前，该企业的区域及规划组级尺码网格需人工构建与管理，导致尺码制定、商品及规划团队的工作负担过重。本研究旨在回答两个研究问题："下一季各尺码网格名称下应为规划人员提供哪些尺码？"及"下一季各规划组应采用哪些尺码？"。我们采用分类模型（机器学习中常见模型之一）解决该问题。通过此研究，运用机器学习技术创建了更自动化的流程，实施后有望降低企业团队的工作负担。与时尚服装行业现有诸多研究不同，本研究聚焦于以库存量单位代表特定尺寸产品的尺码维度。