1. Mechanistic Data Science (MDS) for STEM Education and Applications
Date: Sunday, September 26, 2021
Course format: Hybrid (virtual and in-person at Hyatt Regency Mission Bay)
Instructors: Prof. Wing Kam Liu (Northwestern University), Prof. Zhengtao Gan (Northwestern University), Prof. Mark Fleming (Northwestern University/Fusion Engineering), Prof. Jack Chessa (UTEP) along with postdocs and PhD graduate students for the hand-on experiences.
Course description: This unique seven-hour course is specially designed for students, educators, and researchers interested in the rapidly evolving fields of mechanistic data science and data-driven applications in science, engineering, and technology, machine learning (ML) and artificial intelligence (AI). It will provide a broad perspective on coupling data science tools with mathematical scientific principles to solve intractable problems. Daily-life examples will be used to demonstrate key concepts.
The data science revolution is increasingly being woven into our daily lives and promises to dramatically change the way we will live in the future. Self-driving cars, intelligent gaming, health care, digital media and entertainment, smart homes, smartphone apps and technology, digital personal assistants, personalized shopping, virtual and augmented reality systems, medical drug development, entertainment, internet experiences – these are a few examples of artificial intelligence, machine learning, and data science are already present in our daily lives.
The seven one-hour lectures are:
Introduction to Mechanistic Data Science (MDS). Utilizes data science tools, domain knowledge and scientific principles to identify mechanism; accelerate the science discovery in systematic way; systematic six-steps approach.
Multimodal and multifidelity data collection and generation, Mechanistic feature extraction
Knowledge driven dimension reduction and reduced-order models
Mechanistic learning through regression and classification
System and design 1: Musical instruments sound conversions and forensic engineering
System and design 2: Composite material systems design, rainfall-slope stability
System and Design 3: additive manufacturing, process to property-performance prediction
Reference: Wing Kam Liu, Zhengtao Gan, Mark Fleming, “Mechanistic Data Science for STEM Education and Applications,” Springer, to be published by July 2021.
2. Mechanistic Machine Learning for Engineering and Applied Science
Lecture Instructors: Prof. Steve Sun (Columbia University) and Prof. J.S. Chen (UC San Diego)
Lab Instructors: Nikolaos Vlassis & Bahador Bahmani (Columbia University), Xiaolong He & Kristen Susuki (UC San Diego)
Mechanistic Machine Learning for Engineering and Applied Science will be offered to graduate students and researchers to introduce the practical data analytics, dimension reduction, and machine learning techniques, for a variety of science and engineering applications in materials, structures, and systems.
This course is designed for the audience with a background in mechanics and/or applied physics. The course will provide an overview of four major categories of machine learning techniques (reduced-order methods, geometric learning, manifold learning, and reinforcement learning) and a data-driven model-free framework. Case studies will be used to demonstrate how these learning techniques have enhanced research and technology advancements. These application problems will include a data-driven model-free paradigm for complex material systems, reduced-order modeling of fracture and thermal fatigue analysis, geometric learning for polycrystal and granular systems, and reinforcement learning-enabled multiscale modeling for solid and fluid mechanics problems. Lecture materials and lab handouts will be provided before the short course.
Target Groups: Graduate students, researchers with an understanding of continuum mechanics. Participants must bring their own laptops for the two lab sessions. A course website will be set up for course materials and sample codes repository prior to the short course date.
Topics covered (see attached schedule):
Reduced-order modeling for fracture and thermal fatigue problems
Manifold learning enhanced data-driven modeling of nonlinear materials
Dimension reduction by manifold learning and autoencoders
Geometric learning for predicting material properties
Physics-informed techniques and constraint enforcements
Reinforcement learning for graph model discovery
Copyright (C) Regents of the University of CaliforniaAll Rights Reserved