Functional Data Analysis Group @ UCD

Functional Data Analysis Group @ UCD

We are based in the School of Mathematics and Statistics in the College of Science at University College Dublin.

Former Postdoctoral Researcher

Dr Nurtaj Hossain (current role Postdoc, University of Southern California, Los Angeles)

Research Area: Data-Driven Discovery of Stochastic Differential Equations

Stochastic differential equations (SDEs) are mathematical models that are widely used to describe complex processes or phenomena perturbed by random noise from different sources. The identification of SDEs governing a system is a challenging problem because of the inherent strong stochasticity of data and the complexity of the system’s dynamics.

Current PhD Students

Sajal Kaur Minhas

PhD Project: High-Dimensional Functional Data Analysis

This research is inspired by the ADAPT (Anthropology, DNA, and the Appearance and Perception of Traits) project, which involves analyzing high-resolution 3D facial images. Figure 1 presents the 3D facial structure of an individual's face. The objective of this project is to elucidate the genetic underpinnings of human facial features and enhance our understanding of the ancestry associated with various facial traits.

The project advances statistical methodologies for analyzing complex, high-dimensional data, employing techniques from Functional Data Analysis (FDA). The research focuses on developing novel penalized regression methods to accurately compute 3D surfaces from 3D noisy point clouds.

Figure 1: The smooth manifold representing a subject's 3D facial structure

Uche Mbaka

PhD Project: Fragmented Functional Data Analysis for wired and wireless sensors

Sensors frequently encounter missing segments due to temporary malfunctions, inaccessible sampling sites, or data suppression for confidentiality. These "fragmented" data, illustrated in Figure 2 where dashed lines represent actual curves and dots signify observed fragments, pose significant challenges in analysis.

Functional Data Analysis (FDA) offers statistical methods adept at handling such large-scale and complex datasets by treating data as continuous functions or realizations of a continuous process, with each curve representing a single observation. This project introduces a novel methodology aimed at reconstructing the missing portions of a function based on the observed segments and explores strategies for classifying fragmented functional data.

The initiative also aims to enhance diagnostic capabilities for managing priority diseases in dairy herds. Specifically, it seeks to develop a comprehensive suite of statistical and artificial intelligence models to detect and mitigate diseases like mastitis, leveraging data from on-farm sensors.


Catherine Higgins

PhD Project: Dynamic gene regulatory network construction from high-throughput time course data with application to chronic obstructive lung disease.

One of the fundamental goals in computational systems biology is to model gene expression levels to predict their behaviour under various external/internal conditions. Thanks to advances in technology and collaborative efforts in the scientific community, there is now an abundance of gene expression data collected over time (see: http://genestudy.org/). Identifying gene regulatory network (GRN) models from these big data sets can shed light on the genetic mechanisms that occur when the cellular processes are impaired, for instance, when they are subject to disease or infection.

A network is simply a collection of connected objects. We refer to the objects as nodes and the connections between the nodes as edges. GRNs consider each gene as a node and the direct or indirect interactions between genes as edges. Edges indicate whether the dependence between any pair of genes is inductive, with an increase in the amount of one gene leading to an increase in the other, or inhibitory, with an increase in one leading to a decrease in the other.

An individual’s response to a disease is a dynamic process, which is regulated by this intricate network of many genes and their products. Discovering this dynamic GRN provides new opportunities to identify potential biomarkers and to aid the development of new treatments.


Thiago Americo Da Silva Cardoso

PhD Project: Functional Data Analysis in Complicated 3D Domains

Medical imaging techniques like Functional Magnetic Resonance Imaging (fMRI) have revolutionized modern medicine by offering a non-invasive means to explore the intricacies of the human brain. Recent advancements in medical software enable the transformation of fMRI scans into detailed three-dimensional digital brain meshes, incorporating anatomical measurements. Analyzing such data is complex due to the brain's intricate non-Euclidean geometry and the presence of noise from the imaging process.

We introduce an innovative method for extracting a smooth signal from noisy data associated with the surfaces of 3D objects. This method combines principles from deep learning and Functional Data Analysis (FDA) to create physics-informed neural networks.


Blerta Begu

PhD project: Geo-Spatial Functional Data Analysis

In this research, we focus on space-time point processes and investigate their continuous evolution in space and time. Our contribution lies in introducing an innovative nonparametric methodology for estimating space-time density and intensity. This approach combines maximum likelihood estimation with roughness penalties, employing differential operators across the spatial and temporal domains of interest. The density may be observed over planar or curved domains with intricate geometries, accommodating geographical constraints like complex shorelines in coastal regions or curved areas with intricate topography.

Former MSc Students

  • Jamie Kennedy Project: Spatio-temporal modelling with INLA an application to facial reconstruction
  • Bronagh McCann. Project: A shiny app for gait classification that can determine the severity of Cerebral Palsy in children.
  • Jingfeng Zhu. Project: Forecasting major currency exchange rates
  • Uche Mbaka. Project: Dynamic Analysis of Milk Yield Time Series: A Complex Network Approach
  • Sean Mc Mahon. Project: What factors or farm practices affect the susceptibility of dairy cows to sub-clinical vs clinical mastitis?
  • Shubbham Gupta. Project: What factors or farm practices affect the time between calving and developing mastitis?
  • Akanseoluwa Stephen Adegoke. Project: Exchange Rate Predictability
  • Alessandro Franchini. Project: Modelling oceanographic measurements and streams via spatial regression with differential penalization
  • Clelia Bambini. Project: Spatial Regression Models with Differential Regularization: problems and solutions in Big Data settings
  • Kevin Gallagher. Project: Using Machine Learning methods to analyse the metabolic profiles of professional athletes
  • Mark Dignam. Project: Minimising Errors in Wind Generation Forecasts Using Functional Regression
  • Prajwal Shetty. Project: A Shiny App for Probabilistic Wind Speed Forecasting using Functional Data Analysis
  • Edward Donlon. Project: Uncertainty Quanti cation for Partial Differential Equations with Applications in Geospatial Data Analysis