Bayesian Machine Learning on the Limit Order Book

限价订单簿上的贝叶斯机器学习

• 批准号: 103646
• 金额: $ 88.84万
• 依托单位: BMLL TECHNOLOGIES LIMITED
• 依托单位国家: 英国
• 项目类别: Collaborative R&D
• 财政年份: 2017
• 资助国家: 英国
• 起止时间: 2017 至 无数据
• 项目状态: 已结题
• 来源: https://gtr.ukri.org/projects?ref=103646
• 关键词: Bayesian; Machine; Learning; Limit; Order

BMLL Technologies is a spin-off from the Engineering department at Cambridge University. BMLL works in conjunction with the world's leading trading venues to collate market data at the microscopic level, normalize it and then enable users to extract value from it by combing it with cloud computing power and the latest techniques in machine learning. In 2016 financial data and analytics was a $26.5 billion market (source Burton Taylor Consultancy). This is a number forecast to grow with the importance of empirical survellience and compliance in the financial markets due to new post credit-crunch legislation. Yet many institutions do not have the financial resources or abilities to access or use microscopic data in-house, meaning the current system has been open to accusations of abuse from the high-frequency trading community as popularly related in the best-selling book "Flash Boys". With the InnovateUK funding BMLL will be able to bring to market a software suite allowing pattern recognition to be applied to this data, enabling market abuse to be detected on a global basis for the first time. Subsequently the IP developed for this project will be able to be used in other parts of the BMLL platform, bringing additional benefits.

BMLL技术是从剑桥大学工程系分拆出来的。BMLL与世界领先的交易场所合作，在微观层面整理市场数据，将其规范化，然后通过将其与云计算能力和机器学习的最新技术相结合，使用户能够从中提取价值。2016年，金融数据和分析是一个265亿美元的市场（来源Burton Taylor Consultancy）。这是一个数字预测增长的重要性，实证监督和遵守金融市场由于新的后信贷紧缩立法。然而，许多机构没有财政资源或能力在内部访问或使用微观数据，这意味着目前的系统已被公开的指控滥用从高频交易社区流行的畅销书中有关“闪光男孩”。有了InnovateUK的资助，BMLL将能够向市场推出一个软件套件，允许将模式识别应用于这些数据，从而首次在全球范围内检测市场滥用。随后，为该项目开发的IP将能够用于BMLL平台的其他部分，带来额外的好处。