Hands-on Machine Learning for Political Scientists

What it is about

You will gain practice-oriented knowledge on deep learning and machine-learning methods with reference to the social and political sciences. In the end you can evaluate different ML algorithms and neural network architectures, use deep-learning frameworks such as TensorFlow and PyTorch, prepare data for deep-learning tasks, and implement simple ML projects.

What you will be able to do

• Understand the advantages and disadvantages of various machine-learning algorithms
• Explain architectures of neural networks (e.g. CNN, RNN) and backpropagation
• Apply TensorFlow and PyTorch practically
• Master data preparation and handling training/validation/test data
• Apply central mathematical concepts (linear algebra, optimization, probability theory)
• Link to advanced topics such as computer vision and NLP
• Classify theoretical concepts for social science research (explainability, bias)
• Implement practical machine-learning projects

What the module consists of

• VorlesungConveying the fundamental concepts of machine learning and deep learning
• ÜbungPractical application of lecture knowledge, discussion of advantages/disadvantages and mathematical foundations

Teaching method

• Praktische ProjekteAcquisition of practical experience in handling data and implementing simple projects
• Theoretische KonzepteDiscussion of explainability, interpretability and bias with reference to social and political science questions
• Interaktives LernenUse of analogies, real-world examples and interactive tools to illustrate complex relationships