About

Professor Michael Ludkovski leads the SCiFI Research Group at UC Santa Barbara, focusing on numerical and control methodologies inspired by applications in quantitative finance and insurance. His research emphasizes the interface between machine learning and probabilistic techniques, bridging Financial Mathematics, Actuarial Science, Applied Probability, Operations Research, and Data Science. A significant aspect of his work involves industry outreach in InsurTech, connecting academic research with actuarial practice, and he organizes an annual UCSB InsurTech Summit. His recent research topics include statistical emulation and active learning for computational stochastic control, Gaussian process models in insurance, multi-population longevity modeling, stochastic games, strategic behavior in energy markets, sequential design of experiments in stochastic simulation, machine learning tools for mortality modeling and risk management, meso-scopic behavior of limit order books, and modeling and control of infectious epidemics within stochastic compartmental models.

Research topics

• Computer Science
• Machine Learning
• Mathematics
• Artificial Intelligence
• Algorithm
• Environmental economics
• Reliability engineering
• Engineering
• Business
• Risk analysis (engineering)
• Economics
• Electrical engineering
• Mathematical optimization

Selected publications

• Selecting Critical Scenarios of DER Adoption in Distribution Grids Using Bayesian Optimization

  IEEE Transactions on Power Systems · 2026-01-01

  preprintOpen accessSenior author

  We develop a new methodology to select scenarios of DER adoption most critical for distribution grids. Anticipating risks of future voltage and line flow violations due to additional PV adopters is central for utility investment planning but continues to rely on deterministic or ad hoc scenario selection. We propose a highly efficient search framework based on multi-objective Bayesian Optimization. We treat underlying grid stress metrics as computationally expensive black-box functions, approximated via Gaussian Process surrogates and design an acquisition function based on probability of scenarios being Pareto-critical across a collection of line- and bus-based violation objectives. Our approach provides a statistical guarantee and offers an order of magnitude speed-up relative to a conservative exhaustive search. Case studies on realistic feeders with 200-400 buses demonstrate the effectiveness and accuracy of our approach.

  PublisherOA PDFDOI

• Selecting Critical Scenarios of DER Adoption in Distribution Grids Using Bayesian Optimization

  IEEE Transactions on Power Systems · 2026-01-01

  articleSenior author

  We develop a new methodology to select scenarios of DER adoption most critical for distribution grids. Anticipating risks of future voltage and line flow violations due to additional PV adopters is central for utility investment planning but continues to rely on deterministic or ad hoc scenario selection. We propose a highly efficient search framework based on multi-objective Bayesian Optimization. We treat underlying grid stress metrics as computationally expensive black-box functions, approximated via Gaussian Process surrogates and design an acquisition function based on probability of scenarios being Pareto-critical across a collection of line- and bus-based violation objectives. Our approach provides a statistical guarantee and offers an order of magnitude speed-up relative to a conservative exhaustive search. Case studies on realistic feeders with 200-400 buses demonstrate the effectiveness and accuracy of our approach.

  PublisherDOI

• Analyzing pension fund mortality with Gaussian processes in a subpopulation framework

  Annals of Actuarial Science · 2026-05-08

  preprintOpen accessCorresponding

  Abstract Pension fund populations often have mortality experiences that are substantially different from the national benchmark. In a motivating case study of Brazilian corporate pension funds, pensioners are observed to have mortality that is 40–55% below the national average, due to the underlying socioeconomic disparities. Direct analysis of a pension fund population is challenging due to very sparse data, with age-specific annual death counts often in low single digits. We design and study a collection of stochastic subpopulation frameworks that coherently capture and project pensioner mortality rates via deflator factors relative to a reference population. Superseding parametric approaches, we propose Gaussian process (GP)-based models that flexibly estimate age- and/or year-specific deflators. We demonstrate that the GP models achieve better goodness of fit and uncertainty quantification. Our models are illustrated on two Brazilian pension funds in the context of exogenous national mortality tables. The GP models are implemented in R Stan using a fully Bayesian approach and take into account over-dispersion relative to the Poisson likelihood.

  PublisherOA PDFDOI

• Gaussian process models in actuarial science

  Annals of Actuarial Science · 2026-02-26

  articleOpen access1st authorCorresponding

  Abstract Gaussian Process (GP) modeling is a probabilistic, non-parametric framework for describing spatio-temporal dependence that is well-suited for fitting risk-related surfaces. I summarize the main emerging actuarial use cases of GPs, including their applications in longevity modeling, insurance contract valuation, and loss development. The editorial also discusses further contexts with potential for GP-based approaches.

  PublisherOA PDFDOI

• Stochastic Control

  SpringerBriefs in quantitative finance · 2025-01-01

  book-chapter1st authorCorresponding
  PublisherDOI

• DeepPAAC: A New Deep Galerkin Method for Principal-Agent Problems

  ArXiv.org · 2025-11-06

  preprintOpen access1st authorCorresponding

  We consider numerical resolution of principal-agent (PA) problems in continuous time. We formulate a generic PA model with continuous and lump payments and a multi-dimensional strategy of the agent. To tackle the resulting Hamilton-Jacobi-Bellman equation with an implicit Hamiltonian we develop a novel deep learning method: the Deep Principal-Agent Actor Critic (DeepPAAC) Actor-Critic algorithm. DeepPAAC is able to handle multi-dimensional states and controls, as well as constraints. We investigate the role of the neural network architecture, training designs, loss functions, etc. on the convergence of the solver, presenting five different case studies.

  PublisherOA PDFDOI

• Covariance Kernels

  SpringerBriefs in quantitative finance · 2025-01-01

  book-chapter1st authorCorresponding
  PublisherDOI

• Option Pricing and Sensitivities

  SpringerBriefs in quantitative finance · 2025-01-01

  book-chapter1st authorCorresponding
  PublisherDOI

• Gaussian Processes for Statistical Learning in Actuarial Science

  Foundations for undergraduate research in mathematics · 2025-07-17

  book-chapter1st authorCorresponding
  PublisherDOI

• Gaussian Process Models for Quantitative Finance

  SpringerBriefs in quantitative finance · 2025-01-01 · 6 citations

  bookOpen access1st authorCorresponding
  PublisherOA PDFDOI

Recent grants

• Collaborative Research: ATD: Sequential Quickest Detection and Identification of Multiple Co-dependent Epidemic Outbreaks

  NSF · $213k · 2012–2016

• CDS&E-MSS/Collaborative Research: Sequential Design for Stochastic Control: Active Learning of Optimal Policies

  NSF · $218k · 2015–2019

• Collaborative Research: Gaussian Process Frameworks for Modeling and Control of Stochastic Systems

  NSF · $150k · 2018–2022

• AMPS: Collaborative Research: Stochastic Modeling of the Power Grid

  NSF · $180k · 2017–2022

Frequent coauthors

• Erhan Bayraktar

  Ceva Animal Health (United Kingdom)

  14 shared

• Jimmy Risk

  Pomona College

  12 shared

• René Aïd

  10 shared

• Liangchen Li

  University of Science and Technology of China

  9 shared

• Mickaël Binois

  9 shared

• Ronnie Sircar

  8 shared

• Kyle Bechler

  8 shared

• Robert B. Gramacy

  Virginia Tech

  8 shared

Labs

• SCiFI Research GroupPI

  Simulation and Control in Finance and Insurance