REU Site: Data-driven Materials Design

非技术描述：这个REU网站的目标是培养不同群体的早期本科学生，使他们具备采用数据驱动方法进行材料设计所需的跨学科技能。REU的学生正在学习使用先进的数据分析，如机器学习和人工智能，来有效地识别材料特性的趋势，这些趋势可用于预测未发现材料的特性。这种相对较新的材料信息学方法有望提高效率和生产力，但也需要专门的培训。这个网站提供了一个全面的夏季培训经验，包括：数据和材料科学基础的入门交叉培训，实践研究培训，正式的教育研讨会和讲习班。该网站的重点是为那些通常没有接触到广泛科学资源的学生提供这些机会。除了广泛的支持和指导外，所有参与者都可以通过定制的专业发展和网络机会为工业，学术界和国家实验室的材料科学职业做好准备。技术描述：越来越多地依赖大数据来指导创新，这就需要改变材料科学专业学生的培训方式。对于学生来说，拥有数据驱动方法的经验和熟悉该领域常见的技术已经变得至关重要，然而这些机会对于本科生来说非常有限。这个网站代表了进入这种新培训方法的切入点，并为学生准备了新兴的材料信息学领域。培训将包括基本主题的介绍，如材料科学应用，数据驱动方法的介绍，Python机器学习的介绍，以及实验室技能和安全培训。所有的研究课题本质上都是跨学科的，因为学生们与专门研究不同材料科学方法的初级和次级导师配对。学生探索的科学课题可分为四类，每一类都具有固有的材料信息学成分：(1)材料发现和设计；(2)高通量勘探；(3)定量构性关系；(4)探索和衔接长度尺度。综上所述，这些主题解决了新材料设计和发现、紧急特性和加工策略显著加速的社会需求。减少发现和开发新材料和新技术所需的成本和时间可以对我们的经济、环境和未来产生积极影响。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

NON-TECHNICAL DESCRIPTION: The objective of this REU Site is to train a diverse group of early undergraduate students in the interdisciplinary skills necessary to employ a data-driven approach to materials design. REU students are learning to use advanced data analytics, such as machine learning and artificial intelligence, to efficiently identify trends in materials properties that can be used to predict the characteristics of undiscovered materials. This relatively new materials informatics approach promises to enhance efficiency and productivity, but also requires specialized training. This Site provides a comprehensive summer training experience that includes: introductory cross-training in data and materials science fundamentals, hands-on research training, formal educational seminars, and workshops. The Site focuses on providing these opportunities to students that typically do not have exposure to extensive scientific resources. All participants are prepared for careers in materials science in industry, academia, and national laboratories through customized professional development and networking opportunities in addition to extensive support and guidance. TECHNICAL DESCRIPTION: The increasing reliance on big data to guide innovation necessitates a change in the way that materials science students are trained. It has become essential for students to have experience in data-driven approaches and familiarity with the techniques that are now common to the field, yet these opportunities are very limited for undergraduates. This Site represents an entry point into this new approach to training, and prepares students for the emerging field of materials informatics. Training will include an introduction to essential topics such as material science applications, an introduction to data driven approaches, an introduction to machine learning with Python, and laboratory skill and safety training. All research topics are inherently interdisciplinary since the students are paired with primary and secondary mentors that specialize in different approaches to materials science. Scientific topics explored by the students can be divided into four categories, each with an inherent materials informatics component: (1) materials discovery and design; (2) high-throughput exploration; (3) quantitative structure–property relations; and (4) exploring and linking length scales. Taken together, these topics address a societal need for significant acceleration of new materials design and discovery, emergent properties and processing strategies. Reducing the cost and time it takes to discover and develop new materials and technologies can positively impact our economy, our environment, and our future.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.