Machine Learning - Selected Masterpieces: Actuarial Data Science: A Closer Look at Feature Generation from Unstructured Data
ML models are often referred to as "black boxes", especially so if they are derived from Neural Net-type approaches. But truth be told, many approaches are powerful and do indeed outperform "classical" statistical inference. Consequently, transparency and comprehension are justifiably called upon.
But not last due to the great potential of new ML techniques, there is a tendency to "trusting the algorithm" and (consequently) ignoring the vital importance of curated input data in order to get out what is expected, whereas understanding and shaping the input features is the most important part of the modelling - because that's where one can actually influence the performance the most. In this web session, they will take a closer look at semantically formatting and generating the input space of Machine Learning tools.
As most Actuaries are well-versed in the treatment of highly structured data sets, this web session will look at unstructured data such as time-series (from e.g. Telematics) and fragmented contextuals (like social networks).
The web session is addressed to interested persons who would like to get a first overview and a selective insight into data engineering for Actuarial Data Science. We do not require any specific previous knowledge and start from a very basic level.
In this web session, they will give an overview of the field of Data Engineering for model development. The selective focus of this particular occasion is to broaden the practitioner's understanding of working with unstructured and contextual data sets from both a programmatic and pragmatic perspective.
Please check with your IT department if your firewall and computer settings support web session participation (the programme GoToTraining is used for the web session).
Test Session: On September 1st 2020, 8.00 - 8.30 am GMT, there will be a test session offered to all registered participants to test the software.
The registration fee for each webinar is € 100.00 plus 16 % VAT only. You can book each session individually or register for all three sessions with a special discount of 20%.
Topics:
- What Data Engineering is and why almost no one ever talks about it
- Types of data
- Types of context
- Model enrichment
- Semantic clustering
- Time Series
- Model inception and pipelined design
- Final remarks
Dr Fabian Transchel is a Professor for Data Science at the Harz University of Applied Sciences, Department AI, where his research focus is Insurance and Finance. He is active in the development of the new Actuarial Data Science subjects at the German Actuarial Association (DAV) and a lecturer for the German Actuarial Academy (DAA).