Earlier learning Modules (2022)

06-Dec

Speaker: Ryley McConkey (University of Waterloo)

Learning objectives

Describe the RANS turbulence closure problem Understand the machine learning for turbulence modelling dataset
Explain a typical machine learning data pipeline
Learn how to build and train a neural network using Keras (an interface for TensorFlow)
Learn how to build and train a gradient boosting model using XGBoost
Tune hyperparameters for a machine learning model

Course materials:

All you need to bring is a laptop, and a Kaggle account (https://www.kaggle.com/)
This course will be taught through notebooks running on Kaggle (in your web browser), and will introduce you to Python packages such as scikit-learn, pandas, keras, xgboost.
I will provide most of the code, but you will be able to play with the code during the workshop
This is a hands-on session. You will train multiple machine learning models in your web browser using Kaggle. Please bring your laptop, and register for a free Kaggle account: https://www.kaggle.com/

Speaker: Giovanni Giustini (University of Manchester)

Learning objectives

To provide a practical introduction to simulation methods for “complex” flows i.e., when Navier-Stokes equations alone are not enough.
To provide a minimal theoretical background to multiphysics modelling of such complex flows.
To demonstrate the use of the OpenFOAM code to simulate one such complex flow - two-phase flow with phase change - as a vehicle to familiarise yourselves with the broader topic.
At the end of the session, an OpenFOAM solver will be distributed that simulates two-phase flow with phase change, and a simple test case, to run on the laptops (if interested).

Course materials