developing and validating customized machine learning and data analytics methods for mining cricket

开发和验证采矿板球的定制机器学习和数据分析方法

• 批准号: 2885595
• 金额: --
• 依托单位: University of Manchester
• 依托单位国家: 英国
• 项目类别: Studentship
• 财政年份: 2023
• 资助国家: 英国
• 起止时间: 2023 至 无数据
• 项目状态: 未结题
• 来源: https://gtr.ukri.org/projects?ref=studentship-2885595
• 关键词: developing; validating; customized; machine; learning

Cricket is considered the world's second most popular sport. The plethora of available cricket data and the development of AI technologies have created a massive demand for cricket data analytics. The applications of AI in the cricket domain have increased dramatically during the last two decades.This PhD project is about developing and validating customized machine learning and data analytics methods for mining cricket-related data to support decision-making in talent nurturing and performance profiling.The project will be done in collaboration with the England and Wales Cricket Board (ECB), which is the national governing body of cricket in England and Wales. ECB will also provide access to several years of detailed national and international cricket match data, (longitudinal) performance data of cricket players, and scouting data. The data available to hand is of huge volume, and hence another question to be answered by the successful student is which of the many datasets to use, from which periods, and at what granularity.This project will focus on two main questions/research gaps:1. What are the driving factors for the development of England's cricket players?Addressing this question will involve using statistical analysis and machine learning (e.g. feature importance, dimensionality reduction, clustering) to explore the demographic and developmental sporting history of different player roles (e.g. pace bowlers, batters and wicket keepers) with differing expertise levels to assess the impact these factors have on the development of England's players. Having an improved understanding (that is backed up by the data) of the driving factors, will be useful for nurturing talent and customize training regimes.2. What data and performance metrics can be used to predict success for different formats of cricket (e.g. Test, ODI, iT20)?Addressing this question will involve using feature engineering and prediction methods to identify performance metrics in the First Class (county) game that predict success at International level in Test, One Day International and T20I) formats for English players (potential to explore overseas players). Ultimately, one may want to influence the driving factors for development (Q1) according to the performance metrics identified in Q2 to provide cricket players with a personalized career path.The successful student would be able to develop the research programme within the scope of the two broad topic areas.The project will be done in collaboration with the England and Wales Cricket Board (ECB), which is the national governing body of cricket in England and Wales.

板球被认为是世界上第二受欢迎的运动。大量可用的板球数据和人工智能技术的发展创造了对板球数据分析的巨大需求。人工智能在板球领域的应用在过去的二十年中急剧增加。这个博士项目是关于开发和验证定制的机器学习和数据分析方法，用于挖掘板球相关数据，以支持人才培养和绩效分析的决策。该项目将与英格兰和威尔士板球委员会（ECB）合作，这是英格兰和威尔士板球的国家管理机构。ECB还将提供多年来详细的国家和国际板球比赛数据，板球运动员的（纵向）表现数据和球探数据。现有的数据量巨大，因此另一个需要成功的学生回答的问题是使用众多数据集中的哪一个，来自哪个时期，以及在什么粒度。这个项目将集中在两个主要问题/研究差距：1.英格兰板球运动员发展的驱动因素是什么？解决这个问题将涉及使用统计分析和机器学习（例如特征重要性，降维，聚类）来探索具有不同专业知识水平的不同球员角色（例如投球手，击球手和守门员）的人口统计和发展运动历史，以评估这些因素对英格兰球员发展的影响。对驱动因素有更好的理解（这是由数据支持的），将有助于培养人才和定制培训制度。2.哪些数据和性能指标可用于预测不同形式的板球（例如Test，ODI，iT 20）的成功？解决这个问题将涉及使用特征工程和预测方法来确定一级（县）游戏中的性能指标，这些指标可以预测英国玩家（潜在的海外玩家）在测试，一天国际和T20 I格式中的国际水平成功。最终，人们可能希望根据第二季度确定的绩效指标影响发展的驱动因素（第一季度），为板球运动员提供个性化的职业道路。成功的学生将能够在两个广泛的主题领域范围内开发研究计划。该项目将与英格兰和威尔士板球委员会（ECB）合作完成，这是英格兰和威尔士板球的国家管理机构。