Regression and GLM Course in Actuarial Science
Online 300 h course in regression and GLMs in actuarial science. 6 modules with ASOP No. 12 (risk classification), ASOP No. 25 (credibility) and R (glm, mgcv, glmnet). Assigned tutor and open enrolment.
Take the controls.
Run a real engineering trade-off from this field. Move the sliders, watch the model react in real time and beat the challenge.
What you will be able to do
Explain the move from the linear model to the generalised linear model.
Fit frequency models using Poisson and negative binomial distributions.
Fit severity models and derive the risk premium.
Apply credibility, random effects and regularisation.
Diagnose and validate models using appropriate metrics.
Use models responsibly and without discrimination.
Build a complete technical tariff in a final project.
Who it's for
- Pricing actuary wishing to specialise in Regression and GLMs in actuarial science.
- Pricing analyst wishing to specialise in Regression and GLMs in actuarial science.
- Actuarial data scientist wishing to specialise in Regression and GLMs in actuarial science.
- Non-life tariff manager wishing to specialise in Regression and GLMs in actuarial science.
- Actuarial model validator wishing to specialise in Regression and GLMs in actuarial science.
- Actuarial consultant wishing to specialise in Regression and GLMs in actuarial science.
- Graduates in engineering or sciences seeking to enter non-life insurers.
What you'll work with
- ASOP No. 12 (risk classification)
- ASOP No. 25 (credibility)
- ASOP No. 56 (modelling)
- Directive 2004/113/EC (equal treatment in goods and services)
- Regulation (EU) 2016/679 (GDPR, Art. 22)
- EIOPA supervisory statement on differential pricing practices (2023)
6 modules · 300 h
1.1Linear regression: assumptions and why they fail in insurance1.2Non-normal response variable: counts, amounts and proportions1.3Components of a GLM: random, systematic and link1.4Exponential family and variance function1.5Canonical link versus chosen link1.6Offset and weight: exposure and number of policies1.7Maximum likelihood estimation and iterative algorithm1.8Interpreting coefficients on the multiplicative scale1.9First frequency model step by step1.10Workshop: fitting and interpreting a Poisson GLM
2.1Poisson with exposure offset2.2Overdispersion: quasi-Poisson and negative binomial2.3Excess zeros and two-part models2.4Categorical variables, reference levels and contrasts2.5Grouping levels with low exposure2.6Continuous variables: polynomials and splines2.7Interactions and when they justify their complexity2.8Diagnostics: deviance and Pearson residuals2.9Variable selection and information criteria2.10Practice: frequency model for a motor portfolio
3.1Gamma and lognormal for the average cost3.2Log link and multiplicative effect3.3Treatment of deductibles and limits3.4Large claims: capping and separate treatment3.5Tweedie model for direct pure premium3.6Comparison of the frequency-severity route and the Tweedie route3.7Claims inflation and time adjustment3.8Validation on a holdout sample3.9Building the rating factor table3.10Practice: risk premium by segment
4.1The problem of segments with low exposure4.2Classical credibility and Bühlmann credibility4.3Mixed models with random effects4.4Regularisation: ridge and lasso penalties4.5Generalised additive models for smooth effects4.6Spatial smoothing for geographic zone4.7Cross-validation and choice of penalty parameter4.8Stability of factors across periods4.9Comparison of approaches on the same data4.10Practice: smoothing a geographic factor
5.1Goodness of fit and graphical tests5.2Lift curves and decile analysis5.3Calibration: predicted versus observed5.4Temporal stability and backtesting5.5Prohibited variables and indirect proxies5.6Impact analysis by customer groups5.7Model documentation for the supervisor5.8Model risk and internal controls5.9Communicating results to the business5.10Practice: validation report for a GLM
6.1From risk premium to commercial premium6.2Loadings, expenses and margin6.3Commercial impact of a tariff change6.4Elasticity and portfolio retention6.5Tariff testing and phased roll-out6.6Monitoring the tariff in production6.7Periodic recalibration and version management6.8Tools and reproducibility of calculation6.9Common mistakes when putting a GLM into production6.10Final project: complete tariff from real data
How you'll learn
icon100 % online training on the SEIUM campus, with downloadable material, recorded classes and an assigned technical tutor. No fixed timetable: students progress at their own pace within the intake, which makes it possible to combine with work.
Each module combines theory, guided reading of the original standard and applied exercises. The working software is R (glm, mgcv, glmnet), Python (statsmodels), SAS, Emblem (WTW), Akur8, in free, academic or evaluation versions. Real-world procedures are studied for fitting Poisson and gamma GLMs, diagnostics with residuals and lift curves, and construction of a technical tariff. Assessment is continuous by module plus the final integrative project.
Build it end to end
Your final project follows the real engineering workflow, step by step, reviewed by your tutor.
Explain the move from the linear model to the generalised linear model.
Fit frequency models using Poisson and negative binomial distributions.
Fit severity models and derive the risk premium.
Apply credibility, random effects and regularisation.
Diagnose and validate models using appropriate metrics.
Where this takes you
Requirements
- No prior degree required.
- Internet connection and a computer capable of running the programme's software.
- Sufficient Spanish to follow technical regulatory texts.
Ready to start?
Send your application today. Our admissions team replies within 72 hours with your study plan and funding options.
Frequently asked questions
The programme totals 300 hours spread over 3 to 5 months. That is between 10 and 15 hours per week on average, compatible with a full-time job.
Work is carried out with R (glm, mgcv, glmnet), Python (statsmodels), SAS, Emblem (WTW), Akur8. Free, academic or evaluation versions are used, and the tutor indicates the open-source alternative where one exists.
A SEIUM own diploma upon passing the continuous assessment and the integrative project. It is in-house training, not an official qualification from the Ministry.
Yes. Modules passed in a course or postgraduate diploma in the same area are credited towards the advanced programme. Check the credit transfer policy when you enrol.
Related programmes
Course in Assessing Civil Liability in Autonomous Driving
Online 300-hour course in assessing civil liability in autonomous driving. 6 modules covering Directive (EU) 2024/2853 (liability for defective products), Real Decreto Legislativo 8/2004 (civil liability…
Course on Adapting Insurance Policies to Emerging Climate Risks
Online 300-hour course on adapting insurance policies to emerging climate risks. 6 modules covering Spanish Insurance Contract Law 50/1980, Directive 2009/138/EC (Solvency II) and the Copernicus…
Course on Reinsurance Contract Models
Online course of 300 h in reinsurance contract models. 6 modules covering Principles of Reinsurance Contract Law (PRICL), Ley 50/1980 (Contrato de Seguro, arts. 77 to…
More in Insurance, Risk & Actuarial Data
All Insurance, Risk & Actuarial Data programmes → · How to apply · Fees & financing · Scholarships