Matematika neživotního pojištění 2 - cvičení

Section outline

• Select section Organizational info

  Organizational info

  Tout replier Tout déplier

  - Practicals take place once a week, every Friday 10:40-12:10, starting on 25.2.2022 in K3. Classes will be conducted in present form, as long as it is possible.

  
  - Supporting materials for the practicals will be shared via this Moodle page soon after the session.
  
  - To successfully pass the practicals, you will have to work out and present your home project at the end of the semester. The task will be mainly to model frequency and severity of claims using appropriate methods based on GLM. Specific instructions and data will be assigned to you during the semester. The solution report (incl. relevant source code and software outputs attached in the appendix) must be uploaded into this Moodle page as one pdf document.
  
  - In case of any questions or troubles, do not hesitate to contact me via e-mail (kriz@karlin.mff.cuni.cz) or in person.

  • Select activity Oznámení

    

    Oznámení Forum
• Select section Simple tariffication methods

  Simple tariffication methods

  Covers practicals on 25.2.2022

  • Select activity Exercises 1

    

    Exercises 1 Fichier
  • Select activity Solutions 1

    

    Solutions 1 Fichier
  • Select activity BaileySimon_Rcode

    

    BaileySimon_Rcode Fichier
  • Select activity TotalMarginalSums_WolframCode

    

    TotalMarginalSums_WolframCode Fichier
  • Select activity LogLinearModel_Rcode

    

    LogLinearModel_Rcode Fichier
• Select section GLM - intro

  GLM - intro

  Covers practicals on  4.3.2022

  Contains exponential dispersion family, link function, maximum-likelihood.

  Note: Exercises 2 and 4 serve as preparation for parameter estimation in GLM (will be done in detail next week), but you should be able to work them out.

  • Select activity Exercises 2

    

    Exercises 2 Fichier
  • Select activity Solutions 2

    

    Solutions 2 Fichier
• Select section GLM - parameter estimation

  GLM - parameter estimation

  Covers practicals on 11.3.2022

  Contains equations for MLE and their numerical solution.

  • Select activity Exercises 3

    

    Exercises 3 Fichier
  • Select activity Exercises 3 - data

    

    Exercises 3 - data Fichier
  • Select activity Solutions 3 - theory

    

    Solutions 3 - theory Fichier
  • Select activity Solutions 3 - R code

    

    Solutions 3 - R code Fichier
  • Select activity Solutions 3 - graph

    

    Solutions 3 - graph Fichier
• Select section GLM - multiplicative tariff structure

  GLM - multiplicative tariff structure

  GLM - specifics of the models for claims frequency and claims severity (with multiplicative structure)

  • Select activity Exercises 4

    

    Exercises 4 Fichier
  • Select activity Exercises 4 - claim severity data

    

    Exercises 4 - claim severity data Fichier
  • Select activity Solutions 4 - theory

    

    Solutions 4 - theory Fichier
  • Select activity Solutions 4 - R code with comments

    

    Solutions 4 - R code with comments Fichier
  • Select activity Solutions 4 - graph (Comparison of model predictions)

    

    Solutions 4 - graph (Comparison of model predictions) Fichier
• Select section GLM - variable reduction analysis

  GLM - variable reduction analysis

  GLM - model diagnosis and variable reduction analysis

  • Select activity Exercises 5

    

    Exercises 5 Fichier
  • Select activity Exercises 5 - claim severity data

    

    Exercises 5 - claim severity data Fichier
  • Select activity Solutions 5 - theory

    

    Solutions 5 - theory Fichier
  • Select activity Solutions 5 - R code

    

    Solutions 5 - R code Fichier
  • Select activity Solutions 5 - comments to R results

    

    Solutions 5 - comments to R results Fichier
• Select section GLM - confidence intervals

  GLM - confidence intervals

  GLM - approximate confidence intervals (based on Fisher information) for model coefficients, risk factors and predicted expectations

  • Select activity Exercises 6

    

    Exercises 6 Fichier
  • Select activity Exercises 6 - claim severity data

    

    Exercises 6 - claim severity data Fichier
  • Select activity Solutions 6 - theory

    

    Solutions 6 - theory Fichier
  • Select activity Solutions 6 - R code

    

    Solutions 6 - R code Fichier
  • Select activity Solutions 6 - comments to R results

    

    Solutions 6 - comments to R results Fichier
• Select section Tariffication by GLM - case study

  Tariffication by GLM - case study

  The case study on the frequency-severity approach to tariffication based on GLM applied to real-world Moped insurance data.

  The commented case study  is available on the website of doc. Pešta:

  https://www2.karlin.mff.cuni.cz/~pesta/NMFM402/freq-sev-glm.html

  • Select activity data

    

    data Fichier
  • Select activity R source code

    

    R source code Fichier
• Select section GLM in loss reserving

  GLM in loss reserving

  GLM in loss reserving - discussion of possibilities to approximate distribution of future claims based on historical claims (arranged into the development triangle) and to determine prediction intervals.  See "exercises" file.

  Assignment of the projects to teams - will be distributed via emails.

  • Select activity Exercises 8 (incl. discussion of solution)

    

    Exercises 8 (incl. discussion of solution) Fichier
• Select section Credibility theory

  Credibility theory

  Non disponible