Advanced Machine Learning Methods#

Generative Adversarial Networks#

Content:
Definition and training of the generator and discriminator of one basic and one conditional generative adversarial network (GAN).

Learning Objectives:
The students know and use different types of neural networks for problems in the areas of image processing, time series processing and generative problems. They also learn and apply different training and regularisation methods for optimisation of neural networks.

Link to the repository:
Check out https://git.rz.tu-bs.de/ifn-public/ki4all/generative-adversarial-networks or

git clone https://git.rz.tu-bs.de/ifn-public/ki4all/generative-adversarial-networks

Previous Microcredits: None

Extent: 1.0 ECTS
Responsible: IFN, TU BS

Machine Learning in Network Biology#

Content:
Network biology is an active area of research that covers theory and applications of networks to study complex systems. Many features of biological systems can be represented as networks representing functional or physical relationships between molecular entities, such as genes, proteins, or metabolites. The microcredit course should be useful for students who aim to apply network-based analyses on biological datasets and contains links to the lecture videos and the accompanying course slides for an introduction to biological networks, the network inference: Genie3, disease module mining using GrandForest and patient stratification & biclustering using BiCoN.

Learning Objectives:
This course familiarizes students with machine learning methods for inferring genome-scale biological networks as well as network-based algorithms for analyzing high-dimensional "omics" data.

Link to the repository:
Check out https://git.rz.tu-bs.de/scibiome/ki4all/machine-learning-in-network-biology.git or

git clone https://git.rz.tu-bs.de/scibiome/ki4all/machine-learning-in-network-biology.git

Previous Microcredits: Machine Learning Introduction

Extent: 1 ECTS
Responsible: Peter L. Reichertz Institute for Medical Informatics, TU Braunschweig

Machine Learning Applications in Biomedicine#

Content:
This course allows students to acquire practical experience in implementing ML applications for biomedical problems.

Learning Objectives:
They will learn to conduct and evaluate smaller ML projects on their own, while being supervised by our team.

Link to the repository:
Check out https://git.rz.tu-bs.de/scibiome/ki4all/machine-learning-applications-in-biomedicine.git or

git clone https://git.rz.tu-bs.de/scibiome/ki4all/machine-learning-applications-in-biomedicine.git

Previous Microcredits: Introduction to Artificial Intelligence in Biomedicine

Extent: 1 ECTS
Responsible: Peter L. Reichertz Institute for Medical Informatics, TU Braunschweig

Robustness in Machine Learning#

Content:
As machine learning is increasingly applied to noisy datasets, including biomedical data, it has become important that the models we develop are robust to noise. This repository contains links to the lecture videos and the accompanying course slides for robustness in machine learning, handling imbalanced datasets, performance metrics for model evaluation, common cross-validation techniques and permutation testing and bootstrapping methods.

Learning Objectives:
This course familiarizes students with basic concepts related to model evaluation and practical methods for improving the robustness of machine learning models.

Link to the repository:
Check out https://git.rz.tu-bs.de/scibiome/ki4all/robustness-in-machine-learning.git or

git clone https://git.rz.tu-bs.de/scibiome/ki4all/robustness-in-machine-learning.git

Previous Microcredits: Machine Learning Introduction

Extent: 1 ECTS
Responsible: Peter L. Reichertz Institute for Medical Informatics, TU Braunschweig