back to table of content


INFMW  Deep learning Course INF
Lecturers : Term 1
Course Classification : Master Informatik (Winter-Immatrikulation) CH 4
Language : Deutsch Type VÜS 
Type of examination : PL  Credits
Method of evaluation : written examination 120 min 
Requirements :
Cross References :  
Previous knowledges : Foundations of Artificial Intelligence or Foundations of Machine Learning 
Aids and special features :  
Teaching aims : The students know and understand fundamental concepts of deep learning, such as neural networks, activation functions, and optimization techniques. They are familiar with various neural network architectures (e.g., CNNs, RNNs, Transformers) and can explain these and their areas of application. Students are able to implement, train, and evaluate deep learning models using common frameworks (e.g., TensorFlow, PyTorch) and apply different strategies for model improvement (e.g., regularization, hyperparameter tuning). They can design, implement, and present own deep learning projects. Students critically reflect on the ethical and societal implications of deep learning and artificial intelligence. 
Contents :

* Fundamentals of machine learning * Fundamentals of neural networks: (multi-layer) perceptron, backpropagation, SGD * Specialized neural network architectures, e.g., CNNs, RNNs, transformers * Optimization techniques such as Adam, regularization, dropout, batch normalization * Metrics and evaluation of deep learning models * Frameworks, e.g., TensorFlow, Keras, PyTorch * Application areas, e.g., computer vision, natural language processing, anomaly detection, generative AI * Societal implications: bias, explainability, regulation 

Literature : Bishop, Christopher M.: Deep Learning - Foundations and Concepts. Springer, 2024. - ISBN: 978-3-031-45467-7; ISBN: 978-3-031-45468-4 (eBook) Burkov, Andriy: The Hundred-Page Machine Learning Book. 2019. ISBN: 978-1777005474 Goodfellow, Ian ; Bengio, Yoshua ; Courville, Aaron: Deep Learning. MIT Press, 2016. – http://www.deeplearningbook.org  


back to table of content