8 Nov 2022
Machine Learning: More Art than Science?
Machine learning is currently on everybody’s lips and seems to be a hot topic for many companies. But what is the story behind machine learning and does it really help us to make more accurate predictions? Are we able to implement systems with the ability to automatically learn and improve from experience without being explicitly programmed via training? Are we talking about black box models which cannot be controlled? What about regulatory requirements?
…Is machine learning more art than science?
In this web session we want to learn more about the fast-growing field of machine learning and give answers to these questions.
Organised by the EAA - European Actuarial Academy GmbH.
The web session is open to all interested persons. In particular, this session is not limited to actuaries: Practitioners working in (financial) industry as well as students and researchers with quantitative background are welcome to join.
Technical requirements and test sessionPlease check with your IT department if your firewall and computer settings support web session participations (the programme Zoom is used for this online training). Please also make sure that you are joining the web session with a stable internet connection.In case you want to run and adapt the provided Python code, please install an Python environment; Jupyter notebooks will be sent out for testing purposes in advance.
Purpose and Nature
The notion of machine learning is truly omnipresent in these days. In this web session we thus give participants an introduction to this interesting field and show machine learning techniques. We learn how we are able to implement a workflow and give an overview of several methods such as random forest, gradient boosting, and neural networks. A vital step of machine learning algorithms is the development and training phase of models, in particular, in order to build up powerful applications. Do machine learning models have to be black box applications or are we able to bring light into the darkness? This is a vital question from a practical point of view, which we answer in this web session. Companies are concentrating more and more on using machine learning for several applications, and, therefore, it is important to have a closer look at the current regulatory requirements.
By focusing on all these aspects of machine learning, we give practical examples from the insurance and banking industry: we provide Python codes such that participants directly see how machine learning workflows can be implemented.