Use the Tab and Up, Down arrow keys to select menu items.
This course introduces students to robust practices for the quality-assured application of mathematical models in engineering problem solving. Using simple ODE and machine-learning models drawn from Civil Engineering subdisciplines, students will apply a structured approach to model design, testing and utilisation, including key steps of conceptualisation, verification & validation, calibration, prediction, and uncertainty analysis. By the end of this course, students will appreciate the role of physics-driven and machine learning models for solving engineering problems, their limitations, and how model information must be weighed against financial, sociocultural and legal considerations.
Mathematical and numerical modelling are increasingly present in professional engineering practice. This has been supported by the growing availability of modelling software with easy-to-use interfaces and powerful workflows. Furthermore, recent advances in data science, machine learning and AI have enormous implications for the future of the engineering profession. However, there is a risk that these modern tools could be thoughtlessly used as black boxes without a proper understanding of their correct use and limitations. This course is designed to guide you through a principled modelling process, from start to end (problem to solution). Through computer labs and a design project, this course will challenge you to hone new skills of model calibration, prediction, and uncertainty analysis, and become familiar with modern machine learning concepts. Through groupwork, discussion and communication, you will have opportunities to practice complex mental skills, including decision-making, compromise, criticism, and justification.Examples will be drawn from across Civil Engineering subdisciplines, including structural, transport, water, construction management and geotechnical engineering. You will be assigned a project.This course assumes an introductory level understanding of the Python programming language.
The learning objectives for this course are given below. Your grade will be based on your demonstrated level of mastery of the learning outcomes.1 Select and apply key concepts involved in the practice of physics-driven modelling, including model design, implementation, quality assurance, and application under uncertainty. 2 Derive technical insight by performing a physics-driven modelling study for an engineering problem, reflecting critically on the limits of this knowledge in a wider context.3 Select and apply key concepts of data-driven modelling, including data transformation, statistical analysis, cross-validation, and regression, classification & clustering machine learning algorithms.
COSC131 or equivalent or approval by the Head of Department.
ENCN404
Students must attend one activity from each section.
David Dempsey
Alberto Ardid
All communication with the class will be through lectures, Learn and email.
Electronic copies of course materials will be made available through Learn.
The course comprises two sections on physics-driven and data-driven (machine learning) modelling. Each section is delivered as a series of lectures and computer-based labs that deliver their own learning objectives. Here is the approximate time spent on the different course sectionsPhysics-driven models• Design philosophy (2 weeks) • Quality assurance (2 weeks) Large language models (Generative AI) (2 weeks)Data-driven models• From Data Analysis to Data Science (2 weeks)• Machine Learning Algorithms for Prediction (4 weeks)
Domestic fee $1,268.00
International Postgraduate fees
* All fees are inclusive of NZ GST or any equivalent overseas tax, and do not include any programme level discount or additional course-related expenses.
For further information see Civil and Environmental Engineering .