Courses Detail Information
ME4550J – Introduction to Data-driven Engineering Design
Instructors:
Credits:
3
Pre-requisites:
ENGR1000J, ENGR1010J, ME2500J
Description:
This course introduces the fundamental theories and techniques of data-driven engineering design, aiming to extract insights from large datasets to support decision-making in product/system design. Major topics include fundamental theories of engineering design, basic concepts in decision science, game theory and decision-based design, quantitative methods for customer preference modeling, discrete choice analysis, data analysis and visualization, text mining, supervised/unsupervised learning, deep/reinforcement learning, network analysis, model evaluation, validation and improvement.
Course Topics:
1. Engineering design, design ecosystem, design for market systems
2. Decision-making theory, decision-based design, optimization methods
3. Customer preference modeling, Conjoint Analysis, Discrete Choice Analysis
4. Data-driven design theory and methods
5. Data preprocessing, descriptive data analysis, data visualization
6. Text mining, feature extraction from open data, web crawling, natural language processing, and opinion analysis
7. Supervised/unsupervised learning, logistic regression, neural network, Support Vector Machine, clustering
8. Principal Component Analysis, Network Analysis, deep learning/ reinforcement learning in engineering design
9. Model evaluation, validation and prediction
10. Writing technical reports and making technical presentations