Machine learning methods are also becoming more and more important in the insurance industry. A widespread and in practice very successful method of machine learning uses so-called artificial neural networks.
In the two part lecture we would like to give a first rough overview of the structure of an artificial neural network. We don't care about applications or implementations, but above all about the idea and the mathematics behind it.
The first part is about the structure and the basic components of an artificial neural network. Using a simple example, we show how a neural network processes an input and generates an output.
In the second presentation we show how supervised machine learning works with an artificial neural network. In concrete terms, the so-called backpropagation is explained from a mathematical perspective.