Recent Posts

Player form and Gaussian processes. Part 1: Overview

TLDR When is a player in form (over performing, or enjoying a hot streak) and how long does this last? If there is such an effect, I suspect it will be a result of some complicated system of personal circumstances.
Jan 19, 2022 8 min read Bayesian, Football, Gaussian Process

Uncertainty in xG. Part 2: Partial Pooling

TLDR This is part 2 of an article on fitting a Bayesian partial pooling model to predict expected goals. It has the benefits of (a) quantifying aleatory and epistemic uncertainty, and (b) making both group-level (player-specific) and population-level (team-specific) probabilistic predictions.
Jan 7, 2021 10 min read MCMC, Multi-Level Modelling, Partial Pooling, xG, Football

Uncertainty in xG. Part 1: Overview

TLDR The Expected Goals (xG) metric is now widely recognised as numerical measure of the quality of a goal scoring opportunity in a football (soccer) match. In this article we consider how to deal with uncertainty in predicting xG, and how each players individual abilities can be accounted for.
Dec 10, 2020 7 min read MCMC, Multi-Level Modelling, Partial Pooling, xG, Football

Quantifying the Expected Value of Information

TLDR We sometimes have the option to purchase additional information (like paying for an experiment to be performed, or for a report from an expert consultant) to help us in problems of decision making under uncertainty.
Last updated on Aug 3, 2020 30 min read Bayesian Decision Analysis, Hugin, Uncertainty, VoI, R

Maximum Likelihood Estimation

TLDR Maximum Likelihood Estimation (MLE) is one method of inferring model parameters. This post aims to give an intuitive explanation of MLE, discussing why it is so useful (simplicity and availability in software) as well as where it is limited (point estimates are not as informative as Bayesian estimates, which are also shown for comparison).
Jul 28, 2020 13 min read Parameter Estimation, Maximum Likelihood, R, Python, SciPy, Stan
See all posts

Recent Publications

Quickly discover relevant content by filtering publications.
Domenic Di Francesco, Michael Havbro Faber, Marios Chryssanthopoulos, Ujjwal Bharadwaj
November 2021 In Journal of Data Centric Engineering
Decision-Theoretic Inspection Planning Using Imperfect and Incomplete Data

Decision-Theoretic Inspection Planning Using Imperfect and Incomplete Data

Influence of partial pooling of information on imputation of missing data and expected value of information.
DOI Journal Website
Domenic Di Francesco, Michael Havbro Faber, Marios Chryssanthopoulos, Ujjwal Bharadwaj
April 2021 Special Issue - Decision Making for Extremes
Evaluation of Inspection Features Including Exposure Risk Using a Value of Information Analysis

Evaluation of Inspection Features Including Exposure Risk Using a Value of Information Analysis

Extended Value of Imperfect Information Calculation for Inspection Evaluation.
DOI Journal Website
Domenic Di Francesco, Michael Havbro Faber, Marios Chryssanthopoulos, Ujjwal Bharadwaj
December 2020 Vol. 204C
Consistent and Coherent Treatment of Uncertainties and Dependencies in Fatigue Crack Growth Calculations using Multi-Level Bayesian Models

Consistent and Coherent Treatment of Uncertainties and Dependencies in Fatigue Crack Growth Calculations using Multi-Level Bayesian Models

Bayesian multi-level modelling for partial pooling of information between fatigue crack growth test datasets.
DOI Journal Website
Domenic Di Francesco, Michael Havbro Faber, Marios Chryssanthopoulos, Ujjwal Bharadwaj
August 2020 In OMAE2020
Bayesian Multi-Level Modelling for Improved Prediction of Corrosion Growth Rate

Bayesian Multi-Level Modelling for Improved Prediction of Corrosion Growth Rate

Partial pooling of inspection information from multiple locations to improve probabilistic estimates of corrosion growth rate.
Conference Website
Domenic Di Francesco, Michael Havbro Faber, Marios Chryssanthopoulos, Ujjwal Bharadwaj
December 2019 In IFED2019
Inspection Planning of Hazardous Locations Using a Value of Information Analysis

Inspection Planning of Hazardous Locations Using a Value of Information Analysis

Extension of Bayesian decision theoretic framework to account for risk of inspection activities in quantifying the expected value of information.
Conference Website
Domenic Di Francesco, Michael Havbro Faber, Marios Chryssanthopoulos, Ujjwal Bharadwaj
September 2019 In IPW2019
Bayesian Fatigue Modelling

Bayesian Fatigue Modelling

Quantifying the variability within, and the inter-dependency between all parameters in models that predict fatigue failure of welded steel structures.
Conference website

Research Associate

Alan Turing Institute

Biography

I am a researcher at the Alan Turing Institute and a member of the CSML group at Cambridge University. My research interests include Bayesian statistics and it’s decision theoretic applications, such as quantification of the expected value of information. I also work in football (soccer) analytics.

Interests

  • Bayesian Statistics
  • Information Theory
  • Fracture and Fatigue of Materials
  • Decision Analysis
  • Football Analytics

Popular Topics

Bayesian Bayesian Data Analysis Bayesian Decision Analysis Bayesian Imputation of Missing Data Bayesian Inference Bayesian Modelling Bayesian Statistics Corrosion Growth Decision Analysis Deep Learning Football Markov Chain Monte Carlo Sampling MCMC Multi-Level Modelling Partial Pooling R Stan Uncertainty Value of Information xG

Contact