- Biological and physical foundations of neural networks - Perceptron and delta-rule - Feedforward networks (Multi-layer Perceptrons) and Backpropagation
- Recurrent neural networks and Backpropagation through time
- Hopfield networks and Hebbian learning - Deep neural networks
Exam: written, prerequirements: homework
A comprehensive list of literature will be provided during the lectures Some introductory text books on the topic:
- „Neural Networks - A Systematic Introduction“, Raul Rojas, Springer Berlin 1996 - „Hands-On Machine Learning with Scikit-Learn & TensorFlow“ Aurélien Géron, O’Reilly 2017
Within the course „Artificial Intelligence III: Artificial Neural Networks“, we will cover the basics of artificial neural networks - also called connectionist systems. Such systems, inspired by biological neural networks, are built from a set of simple processing units (neurons) which are connected (through synapses) to process noisy information. During the course, the following types of networks are introduced: - Simple perceptrons - Feed forward neural networks (including auto-encoders) - Recurrent neural networks - Hopfield networks - Some of the recent deep neural networks In addition, we will discuss the mathematics of the corresponding learning algorithms.