6.00pm - 6.30pm: Tea/Coffee Reception
6.30pm - 8.00pm: Meeting
This presentation investigates the use of data science and machine learning techniques in the area of general insurance claim modelling. Specifically we look at whether regularised regression modelling (the lasso) can be used to construct automatically a stochastic model that captures complex claims experience (including superimposed inflation, legislative changes, changing development patterns, etc.), yet still delivers an interpretable model.
Engineering a suitable set of features/basis functions for inclusion in such a model is key to its success and we will discuss some of the important factors in building the set of basis functions. This form of modelling is illustrated by application to both synthetic data sets, with known features, and a real data set with which the authors are familiar. The outcome is the automated modelling of highly complex data sets.
A paper discussing the contents of this presentation is available at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3241906. This presentation is very similar to that given at GIRO 2018, under the same title.