Research areas
Actuarial and insurance mathematics
Stochastic modelling in life and non-life insurance, ruin theory, risk processes, credibility theory, loss distributions, survival models, reserving mathematics, premium principles, dependence modelling, extreme value theory, and mathematical foundations of risk measures.
Quantitative risk modelling and financial econometrics
Time series, Markov-switching models, risk measures, backtesting, regime detection, stochastic processes, portfolio risk, portfolio optimization
Machine learning, XAI and data-driven insurance
Explainable AI, ML in tariff modelling, insurance telematics, graph mining, fraud/risk detection, interpretable predictive systems.
Stochastic and computational modelling
Monte Carlo methods, simulation-based inference, numerical experiments, dynamic systems, agent-based modelling
Papers and manuscripts
Airline passenger satisfaction classification using machine learning
Data Mining course project, June 2026
A binary classification project predicting airline passenger satisfaction from demographic, travel-related and service-quality features. The analysis compares logistic regression, Random Forest, HistGradientBoosting and MLP models, with preprocessing pipelines designed to avoid data leakage and model evaluation based on accuracy, precision, recall, F1 and ROC AUC.
Keywords: data mining, machine learning, classification, airline satisfaction, logistic regression, Random Forest, HistGradientBoosting, MLP, ROC AUC.
Links: PDF / GitHub
Survival analysis of heart failure clinical records
Survival analysis course project, June 2026
A survival analysis project modelling time-to-death in patients with heart failure under right censoring. The study compares nonparametric Kaplan-Meier and Nelson-Aalen estimators, actuarial life tables, parametric AFT models, a Bayesian extension and Cox proportional hazards models with diagnostic assessment.
Keywords: survival analysis, duration analysis, right censoring, Kaplan-Meier, Nelson-Aalen, AFT models, Weibull model, Bayesian survival model, Cox proportional hazards.
Links: PDF / GitHub
Bayesian models for high-income classification
Bayesian econometrics homework assignments, 2026, June 2026
Two Bayesian modelling studies on the Adult Census Income dataset: a conjugate linear probability model and a Bayesian logistic regression estimated with Stan/NUTS. The projects compare analytical and simulation-based Bayesian inference for modelling the probability of earning more than 50K USD.
Keywords: Bayesian econometrics, Adult Income, linear probability model, logistic regression, Normal-Gamma prior, Stan, MCMC, HPDI, Bayes factors.
Links: PDF_1 / PDF_2 / GitHub
Cardinality-constrained Markowitz portfolio optimization using metaheuristics
Course project report, June 2026
A Monte Carlo study of Markowitz portfolio optimization with cardinality constraints, based on synthetic GARCH-generated return scenarios and solved using simulated annealing, genetic algorithms, and particle swarm optimization.
Keywords: Markowitz optimization, cardinality constraints, GARCH, Monte Carlo, genetic algorithm, particle swarm optimization, simulated annealing.
Links: PDF / GitHub
Backtesting investment fund portfolios using GARCH and Markov-switching models
Bachelor thesis, July 2025
A comparative study of classical volatility models and regime-switching models for conservative and aggressive investment fund portfolios.
Keywords: GARCH, MS-GARCH, VaR, Expected Shortfall, investment funds, backtesting.
Links: PDF / GitHub