Using R for insurance

Bucharest 22 November 2019

Trainer: Frederick Borgers

Details

The data requirements in insurance are ever growing larger. For Solvency II and other purposes, quantitative and qualitative data requirements are increasing. Usually a database or datawarehouse is available in the company, but adequate tools to analyse the data may lack, resulting in over reliance on Microsoft Excel – Especially if commercial software such as SAS is missing.

We try to overcome this issue by presenting R as a possible tool for making typical insurance analysis. At the same time, well known insurance concepts will be introduced, such as the chain-ladder for reserving and GLM for predictive modelling.

- Prepare a dataset for reserving and then apply chain-ladder methodology in R

- Descriptive statistics of a dataset: histograms, statistical tests, distribution checks

- Using generalized linear models and other regression techniques used for predictive modelling

The focus lies on the practical implementation of a number of concepts in R, rather than the theoretical description of those techniques. Exercises will be given during the course to make the users familiar with R and a small part will be dedicated to the newest techniques in the field.

Target group

  • Junior actuaries
  • More senior actuaries or persons in data/controlling function who are eager to find out about R
  • Students who want to complement their actuarial education with practical insights

Course objectives

  • Make the participants familiar with common insurance technical aspects
  • Allow the participants to start using R in their daily work
  • Give a brief overview of common techniques used for pricing and reserving

Subjects

Theoretical part

- Common statistical techniques used in insurance

  • Data analysis
  • Reserving
  • Pricing

Practical part

- Introduction to R

  • How to prepare a dataset for R
  • Import/Export of data in R
  • General rules for coding with R

- Reserving

  • Building a dataset specifically for reserving
  • Chain-Ladder technique
  • Derive a best estimate including measures such as including/excluding link factors

- Pricing

  • Building a GLM model to predict the risk premium starting from an available dataset

Trainer

Frederick Borgers is the head of Group Property & Casualty Pricing at UNIQA Insurance Group AG. Frederick was an actuary within the International Actuarial Specialist Programme at Eureko, a non-life actuary at Uniqa Insurance Group AG and the Actuarial Project Manager at UNIQA Asigurări.

Duration / Period

The workshop will last for 6 hours and will take place on 7th of December 2018, between 09:30 and 16:30.

Discounts

At the above fee the following discounts will be granted:

  •     5% for registration and payment until 1st of November 2018;
  •     5% for the registration of more than 2 persons from the same organization;

    The discounts can be cumulated.

Apply for course
Helpful information to complete the form.




 Newsletter ISF

Abonează-te acum pentru a primi ultimele noutăți. Suntem prezenți și pe rețelele de socializare și