Training "Data-driven Modeling and Simulation"
The basics and methods will be communicated for data-driven modeling and simulation based on physics-informed machine learning with the software OptiY®. Concurrently, the course contains also practical exercises for challenges of CAD/CAE-engineers in the industry.
Course Content
1. Workflow-Management
- Parameters, Criteria and Constraints in Workflow
- Data Import and Export Assistant
2. Data-Analysis
- Parallel Chart
- Scatter-Plot (2D and 3D)
- Histogram and Box-Plot
- Correlation Matrix
- Tables
3. Approximation
- Gaussian Process
- Hilbert Space
- Reproducing Kernal Hibert Space
- Nonlinear Least-Square
- Residual-Plot and Characteristics
- 1D-, 2D- and 3D-graphical Presentation
4. Gradient-based Optimization Methods
- Gradient Descent
- Stochastic Gradient Descent
- L-BFGS
5. Physics-informed Machine Learning
- Partial Differential Equations
- Boundary or Initial Conditions
- Constraints
- Parameters
- Automatic Differentiation
6. Applications
- System Simulation of a double Spring Mass Damper
- Material Modeling of Stress-Strain-Characteristics
- Parameter Identification of Partial Differential Equations
Details
Referent:
Dr. The-Quan Pham
Time:
1 Day from 9.00 to 17.00
Place:
Würzburg, Germany
Cost:
500,- Euro (excluding VAT)
It includes training material, drinks and lunch
Discount:
40% for University Members
20% for Registration with multiple Persons
Date: available on request
