Postgraduate Diploma in Predictive Claims Models with Machine Learning
Online Postgraduate Diploma of 800 h in Predictive Claims Models with Machine Learning. 8 modules covering Directive 2009/138/CE (Solvency II), Regulation (EU) 2024/1689 (Artificial Intelligence Act) and Python (scikit-learn, LightGBM). Assigned tutor and open enrolment.
Take the controls.
Run a real engineering trade-off from this programme. Move the sliders, watch the model react in real time and beat the challenge.
What you will be able to do
Prepare and clean claims databases for modelling.
Build benchmark statistical models for frequency and severity.
Apply machine learning algorithms to claims prediction.
Model large claims and their impact on the portfolio.
Use predictive models to prioritise, allocate and reserve claims.
Interpret and validate models using explainability and backtesting techniques.
Deploy and monitor models in production under governance criteria.
Who it's for
- Claims data scientist who wants to specialise in Predictive Claims Models with Machine Learning.
- Reserving actuary who wants to specialise in Predictive Claims Models with Machine Learning.
- Machine learning engineer in insurance who wants to specialise in Predictive Claims Models with Machine Learning.
- Claims analyst who wants to specialise in Predictive Claims Models with Machine Learning.
- Model validation manager who wants to specialise in Predictive Claims Models with Machine Learning.
- Advanced analytics consultant in insurance who wants to specialise in Predictive Claims Models with Machine Learning.
- Graduates in engineering or science seeking to enter non-life insurers.
What you'll work with
- Directive 2009/138/CE (Solvency II)
- Regulation (EU) 2024/1689 (Artificial Intelligence Act)
- Regulation (EU) 2016/679 (GDPR)
- ISO/IEC 23053 (framework for AI systems using machine learning)
- ISO/IEC 23894 (AI risk management)
- ISO/IEC TR 24027 (bias in AI systems)
- Delegated Regulation (EU) 2015/35 (data quality in Solvency II)
8 modules · 800 h
1.1Structure of policy and claims databases1.2Exposure and accident year1.3Frequency, severity and average cost1.4IBNR claims and development1.5Actuarial data quality1.6Risk variables by line of business1.7Large claims and truncation1.8Inflation and amount adjustment1.9External data for enrichment1.10Case study: exploratory analysis of a motor portfolio
2.1Poisson and negative binomial GLM2.2Gamma and lognormal severity models2.3Tweedie distribution for pure premium2.4Zero-inflated models2.5GAM and non-linear effects2.6Exposure offset2.7Variable selection and regularisation2.8Residual diagnostics2.9Interpretation of relativities2.10Practical: frequency GLM in R or Python
3.1Decision trees and Random Forest3.2Gradient boosting with Poisson objective3.3LightGBM and CatBoost with categorical variables3.4Neural networks and CANN3.5Survival models for time to claim3.6Hyperparameter tuning3.7Temporal cross-validation3.8Metrics: deviance, Gini and lift3.9Comparison with GLM3.10Lab: frequency GBM versus GLM
4.1Heavy-tailed distributions4.2Extreme value theory4.3Thresholds for large claims4.4Quantile regression4.5Mixed severity models4.6Prediction of bodily injury claims4.7Ultimate cost of open claims4.8Individual case reserve models4.9Prediction uncertainty4.10Practical: severity model with Pareto tail
5.1Triage at first notification5.2Cost prediction at opening5.3Detection of complex claims5.4Litigation prediction5.5Claims handler allocation5.6Recoveries and subrogation5.7Claims volume forecasting5.8Catastrophic claims and peaks5.9Integration with claims handling systems5.10Case study: bodily injury triage model
6.1Permutation feature importance6.2SHAP and partial dependence6.3Interpretable surrogate models6.4Independent model validation6.5Stability and backtesting6.6Bias and fairness in prediction6.7Solvency II requirements on data and models6.8European AI Regulation in insurance6.9Model documentation6.10Workshop: explainability report for business
7.1Reproducible training pipelines7.2Data and model versioning7.3Batch and real-time deployment7.4Drift monitoring7.5Periodic retraining7.6Integration with pricing and reserving7.7Performance dashboards7.8Change management with actuaries7.9Model cost and benefit7.10Practical: retraining pipeline with MLflow
8.1Definition of the business objective8.2Data preparation8.3GLM benchmark model8.4Machine learning models8.5Validation and explainability8.6Economic impact analysis8.7Deployment plan8.8Monitoring plan8.9Report and presentation8.10Final review with the tutor
How you'll learn
100 % online training on the SEIUM campus, with downloadable material, recorded classes and an assigned technical tutor. No fixed timetables: students progress at their own pace within the intake, which makes it possible to combine the programme with work.
Each module combines theory, guided reading of the original standard and applied exercises. The working software is Python (scikit-learn, LightGBM), R (tidymodels), H2O.ai, MLflow, SHAP, in free, academic or evaluation versions. Assignments are based on real procedures for predicting claims frequency and cost on historical portfolios, modelling large claims with heavy-tailed distributions, temporal cross-validation and drift monitoring. Assessment is continuous per 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.
Prepare and clean claims databases for modelling.
Build benchmark statistical models for frequency and severity.
Apply machine learning algorithms to claims prediction.
Model large claims and their impact on the portfolio.
Use predictive models to prioritise, allocate and reserve claims.
Where this takes you
Requirements
- Prior technical training or professional experience in the sector. No university degree is required.
- Internet connection and a computer capable of running the programme's software.
- Sufficient level of 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 800 hours, spread over 6 to 8 months. That is between 10 and 15 hours per week on average, compatible with a full-time job.
You work with Python (scikit-learn, LightGBM), R (tidymodels), H2O.ai, MLflow, SHAP. Free, academic or evaluation versions are used, and the tutor indicates the open-source alternative where one exists.
A SEIUM diploma awarded upon passing the continuous assessment and the integrative project. This is SEIUM's own training, not an official qualification from the Ministry.
Yes. Modules passed in a course or postgraduate diploma in the same area are recognised towards the advanced programme. Check the recognition policy when you enrol.
Related programmes
Postgraduate Diploma in Elasticity and Customer Segmentation in Insurance
Online 800-hour postgraduate diploma in elasticity and Customer Segmentation in Insurance. 8 modules covering Delegated Regulation (EU) 2017/2358 (product governance), Directive (EU) 2016/97 (IDD, customer demands…
Postgraduate Diploma in IoT Integration in the Insurance Ecosystem
Online 800-hour Postgraduate Diploma in IoT Integration in the Insurance Ecosystem. 8 modules covering Regulation (EU) 2024/2847 (cyber resilience), ETSI EN 303 645 (cybersecurity of consumer…
Postgraduate Diploma in Fraud Detection Techniques in Claims
Online 800-hour Postgraduate Diploma in Fraud Detection Techniques in Claims. 8 modules covering Regulation (EU) 2024/1689 (Artificial Intelligence Act), Regulation (EU) 2016/679 (GDPR) and Python (scikit-learn,…
More in Insurance, Risk & Actuarial Data
All Insurance, Risk & Actuarial Data programmes → · How to apply · Fees & financing · Scholarships