Mar 13, 2025  
2023-2024 Undergraduate Academic Catalog-June Update 
    
Catalog Navigation
Catalog Home
Academic Calendar
University Overview
Admissions
Student Financial Services
Academic Regulations and Policies
Degree Programs, Minors, and Certificates
Degrees, Minors, and Certificates by Department
Course Descriptions
The Raider Core
MSOE Honors Program
Global Opportunities
Reserve Officers’ Training Corps. (ROTC)
Administration, Regents, and Advisory Committees
Departments and Faculty
Campus Map
Student Resources
My Portfolio
2023-2024 Undergraduate Academic Catalog-June Update [ARCHIVED CATALOG] Add to Portfolio (opens a new window)

MEC 4678 - Data and Mechanical Engineering

2 lecture hours 2 lab hours 3 credits
Course Description
This course introduces students to data-driven methods, machine learning, and optimization in mechanical engineering contexts. The course starts with a crash course in Python programming and review of linear algebra to support course content, homework assignments, and projects. Students will learn the fundamentals of and use computational tools involving data heavy problems relevant to fluid mechanics, solid mechanics, and control theory. The tools and approaches covered and used in the course will prepare students for jobs and/or graduate studies involving data-oriented problems in mechanical engineering. (prereq: MEC 2020 , MEC 3120 , MEC 4310 ) (quarter system prereq: ME 2003, ME 3104, ME 4302)
Course Learning Outcomes
Upon successful completion of this course, the student will be able to:
  • Identify and apply appropriate dimensionality reduction to complex data
  • Identify and apply appropriate regression analysis to a high-dimensional data set
  • Construct and train a neural network for a classification problem
  • Construct and train a neural network for a regression problem
  • Implement a data driven control technique for a complex dynamic system
  • Use a physics informed neural network (PINN) to solve a design optimization problem
  • Use a PINN to solve for unknown coefficients for a partial differential equation
Prerequisites by Topic
  • Dynamics (kinematics and kinetics principles for general plane motion)
  • Fluid mechanics (Conservation of Mass, Navier Stokes equations)
  • Controls (mathematical models of systems, feedback control systems characteristics)
Course Topics
  • Programming for and working with data in mechanical engineering
  • Dimensionality reduction
  • Regression analysis and classification
  • Neural networks
  • Optimal control techniques  
  • Reinforcement learning
  • Physics informed neural networks (PINNs)
Laboratory Topics
  • Programming and data-oriented library crash course
  • Dimensionality reduction of experimental data
  • Classification of high-dimension data 
  • Regression analysis of high-dimension data 
  • Identifying and adjusting a problem for an appropriate approach - regression analysis and classification
  • Construction and implementation of neural networks
  • Using neural networks for predictive regression and classification in manufacturing contexts
  • Application of optimal control techniques  
  • Reinforcement learning for controlling complex systems
  • Optimization using physics informed neural networks (PINNs)
  • Using inverse PINN solutions to determine complex physical relationships
Coordinator
Dr. Nathan Patterson


Add to Portfolio (opens a new window)