Mini-project
Insurance
Motor claims statistics
Study how often and how much motor policies claim, and test common beliefs about drivers.
The brief
Underwriters believe young drivers and powerful cars claim more. Test these beliefs with statistics on real motor policy data: claim frequency and severity by driver age, vehicle power and region, with confidence intervals rather than hunches. Good looks like: correct tests, effects with uncertainty, and charts underwriters can use.
What you will build
- Explore claim counts and amounts per policy-year
- Compare frequency by driver age and vehicle groups
- Test differences and compute confidence intervals
- Fit a Poisson model for claim frequency
- Summarise what is and is not supported by the data
Free, open data to build it with
- French motor claims (freMTPL2freq and freMTPL2sev) ↗ Kaggle
- CASdatasets: actuarial datasets ↗ CASdatasets (GitHub)
Check each dataset's licence before you publish your work. Your mentor gives your team its own version of the brief (data, scope or region), so every team's project is its own.