FSM3001 Data-driven Methods in Engineering 7.5 credits

The amount of data generated in the modern engineering solutions is immense, and it will only grow. It is imperative to exploit the data for model development, continuous evaluation of damage and testing. This graduate-level course focuses on state-of-the-art techniques in data-driven modeling. The course introduces relevant aspects of probability theory, optimization, and the basics of machine learning and deep learning. The course covers a variety of modeling and learning methodologies and algorithms, such as modern neural-network architectures, modal decompositions, identification of linear and nonlinear dynamics, and other advanced topics in data-driven modeling. The emphasis will be on the application of modern data-driven modeling tools to solid and fluid mechanics, dynamical systems, control and more applied engineering settings (naval and vehicle engineering, aeronautics, railways and wind energy). The course will finalize by a project where students apply techniques developed in this course (or extensions thereof) to a particular dataset/problem. In this sense, the course is designed to prepare students with the tools and knowledge to use the techniques to conduct novel datadriven engineering, research and development.
Choose semester and course offering
Choose semester and course offering to see current information and more about the course, such as course syllabus, study period, and application information.
Content and learning outcomes
Course contents
Dimensionality Reduction (Part I) This section introduces tools for finding low-dimensional representations of high-dimensional data, which allows for data to be efficiently stored, transferred, and analyzed.
Machine Learning and Data Analysis (Part II) This section will give a relatively brief tour through aspects of data analysis, from classical curve fitting to neural networks and deep learning, building on the material introduced in Part I.
Dynamics, Control and Reduced-Order Models (Part III). In this section, we assume that the data that we are studying comes from some underlying physical laws (in the context of dynamical systems, solid mechanics, fluid mechanics, etc.), which can be learned/approximated from data, or from some combination of data and physics.
Final Project (Part IV).
The students will apply the techniques developed in this course (or extensions thereof) to a dataset/problem of their own choosing.
Intended learning outcomes
After taking the course the students should be able to:
- Understand the meaning and significance of mathematical operations required to process, represent, and approximate data.
- Understand the objectives, advantages, and disadvantages of various data-driven modeling techniques.
- Learn how to load and manipulate large datasets in Matlab and/or Python.
- Develop the required skills to apply various data-driven algorithms to potentially large and complex datasets
- Interpret the results of modeling algorithms to build an enhanced understanding of a given dataset.
- Interpret and understand the physics of the underlying system that the data comes from.
- Be able to assess the implications of the developed data-driven solutions for sutainable development.
Course disposition
Literature and preparations
Specific prerequisites
A good understanding of standard topics in engineering mathematical analysis will be very helpful. In particular, a strong background in linear algebra, differential equations, and optimization will be beneficial. Since hands-on data-driven modeling will invariably require some coding, familiarity with Matlab, Python or other similar languages/platforms will be helpful.
Recommended prerequisites
Equipment
Literature
Examination and completion
If the course is discontinued, students may request to be examined during the following two academic years.
Grading scale
Examination
- PRO1 - Project, 3.5 credits, grading scale: P, F
- TEN1 - Written exam, 2.0 credits, grading scale: P, F
- ÖVN1 - Home work, 2.0 credits, grading scale: P, F
Based on recommendation from KTH’s coordinator for disabilities, the examiner will decide how to adapt an examination for students with documented disability.
The examiner may apply another examination format when re-examining individual students.
Opportunity to complete the requirements via supplementary examination
Opportunity to raise an approved grade via renewed examination
Examiner
Ethical approach
- All members of a group are responsible for the group's work.
- In any assessment, every student shall honestly disclose any help received and sources used.
- In an oral assessment, every student shall be able to present and answer questions about the entire assignment and solution.
Further information
Course web
Further information about the course can be found on the Course web at the link below. Information on the Course web will later be moved to this site.
Course web FSM3001