Reinforcement learning and optimization for mass customization of embroidery

刺绣大规模定制的强化学习与优化

• 批准号: 509935-2017
• 负责人: Yu, JiaYuan
• 金额: $ 2.77万
• 依托单位: Concordia University
• 依托单位国家: 加拿大
• 项目类别: Collaborative Research and Development Grants
• 财政年份: 2017
• 资助国家: 加拿大
• 起止时间: 2017-01-01 至 2018-12-31
• 项目状态: 已结题
• 来源: https://www.nserc-crsng.gc.ca/ase-oro/Details-Detailles_eng.asp?id=618886
• 关键词: Reinforcement; learning; optimization; mass; customization

The objective is to demonstrate applications of machine learning and operations research in the fashion industry. In particular, we solve a problem of mass customization by automatically generating the embroidery machine instructions from images uploaded by customers. Traditionally, these instructions are painstakingly created by trained workers using specialized software. The machine instructions are fed to embroidery machines to create embroideries on selected garments. The proposed system will be the first of its kind to automate the process. Our solution approach uses algorithms for neuro-dynamic programming and traveling salesman problems with clusters.

其目的是展示机器学习和运筹学在时尚业的应用。特别是，我们解决了大规模定制的问题，通过从客户上传的图像自动生成刺绣机指令。传统上，这些指令是由训练有素的工人使用专门软件煞费苦心地创建的。机器指令被送入刺绣机，在选定的服装上制作刺绣。拟议中的系统将是此类系统中第一个将该过程自动化的系统。我们的解决方法使用神经动态规划和带簇的旅行商问题的算法。