Excellence in Research: Research in Machine Learning and Its Application

该项目旨在开发一个数据驱动的机器学习研究计划，并伴有数据科学和机器学习的综合本科教育课程。该研究团队由来自本尼迪克特学院（Benedict College）、历史悠久的黑人学院和大学（HBCU）、南卡罗来纳大学（university of South Carolina）和北卡罗莱纳州立大学（North Carolina State university）的教师组成，他们在优化、控制理论、统计学、应用和计算数学以及工程学方面具有互补的专业知识，将开发机器学习的新方法，并将这些方法应用于各种应用，包括疟疾流行和糖尿病建模；发现一些材料和生命科学问题的本构规律和机制，如心脏组织建模和设计合成离子分离膜；将机器学习模块整合到混合多尺度模型中，用于模拟三维生物制造中各种器官和组织结构的血管生成（血管生成是指在血管系统形成和发展的早期阶段，由原有血管形成新血管的生理过程）。此外，该项目将创建一个创新的培训计划，以教育本科STEM学生，并重组本尼迪克特学院的附属教师，为他们做好数据科学和人工智能相关工作的准备，并在未来使用数据驱动的方法进行研究。计算机和学习实验室将为参与研究的教师和学生提供必要的计算机设备，以进行研究和教育活动。项目团队将专注于使用数据科学和机器学习工具可以解决和改进的几个应用问题：(1)在疟疾流行和糖尿病疾病建模中，使用多目标优化方法来改进机器学习在聚类、特征选择、知识提取和集成生成方面的结果；(2)运用灰色模型改进财务分析预测；(3)建立包括疟疾和其他疾病在内的疾病流行预测模型；(4)发现特定材料的本构规律和机制，以及基于深度神经网络的异质心脏组织和器官的应力-应变本构关系和合成离子分离膜中的离子传输机制等生命科学问题；(5)结合机器和深度学习工具来校准相互作用能量，并加速血管生成混合多尺度模型的蒙特卡罗模拟。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。

This project is aimed at developing a data-driven machine learning research program accompanied by an integrated undergraduate educational curriculum in data science and machine learning. The research team, consisting of faculty from Benedict College, a historically black college and university (HBCU), University of South Carolina, and North Carolina State University with complementary expertise in optimization, control theory, statistics, applied and computational mathematics, and engineering will develop new methods in machine learning and employ these methods in a variety of applications, including modeling malaria epidemics and diabetes; discovering constitutive laws and mechanisms for some selected materials and life science problems such as modeling heart tissues and designing synthetic ion separating membranes; incorporating machine learning modules into a hybrid multiscale model for simulating angiogenesis (angiogenesis is the physiological process through which new blood vessels form from preexisting vessels, formed in the earlier stage of formation and development of vascular system) of various organs and tissue constructs in 3D biofabrication. Additionally this project will create an innovative training program to educate undergraduate STEM students and to retool affiliated faculty in Benedict College to prepare them for data science and artificial intelligence related jobs and for conducting research using data-driven approaches in the future. The computing and learning laboratory will provide the necessary computing facility for participating faculty and students to carry out the research as well as educational activities. The project team will focus on several application problems that can be solved and improved using data science and machine learning tools: (1) using a multi-objective optimization approach to improve machine learning outcome in clustering, feature selection, knowledge extraction, and ensemble generation in modeling malaria epidemics and diabetes disease; (2) using Grey models to improve predictions in financial analysis; (3) developing forecasting models for disease epidemics including malaria and other diseases; (4) discovering constitutive laws and mechanisms in selected materials and life science problems such as stress-strain constitutive relations based on deep neural networks for heterogeneous heart tissues and organs and ion transport mechanisms in synthetic ion separating membranes; (5) coupling machine and deep learning tools to calibrate interaction energies and accelerate Monte Carlo simulations in a hybrid multiscale model for angiogenesis.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.