If you made any changes in Pure these will be visible here soon.

Personal profile

Research profile

Mathematical Statistics is a subfield of statistics where mathematical tools are used to design and study statistical procedures. Within Mathematical Statistics, Paulo’s research focuses on non-parametric (high- or infinite dimensional) models. Such large models are useful to avoid having to make potentially restrictive assumptions. Paulo work mainly with three different types of approaches: penalised estimation (used to reconstruct complex signals from observations), recursive estimation (a tool for online learning), and Bayesian estimation (a powerful tool for statistical learning that allows expert knowledge to be incorporated into the inference in a natural way). 

Currently Paulo focuses on the use of these techniques in medical applications, working jointly with researchers at Philips and other departments of the TU/eHe is also involved with the SMM research program of the Data Science Center Eindhoven (DSC/e) and does external consulting work.

Academic background

Paulo Serra obtained Licenciatura in Applied Mathematics at New University of Lisbon, Portugal, followed by an MSc from Utrecht University, and a PhD from the Eindhoven University of TechnologyHe has been a postdoc at the Korteweg-de Vries Institute, of the University of Amsterdam. Before that, he spent two years as a postdoc at the Institute for Mathematical Stochastics of the University of Göttingen, in Germany. Earlier, he was a junior researcher for a year at CA3 group at UNINOVA, in Portugal, working on project NOMDIS developed for ESA/ESOC. Paulo has contributed to several publications, book chapters, and conference proceedings. Since 2016 Paulo has been an Assistant Professor at the Stochastics group of the Department of Mathematics and Computer Science at Eindhoven University of Technology (TU/e).

Fingerprint Dive into the research topics where Paulo Serra is active. These topic labels come from the works of this person. Together they form a unique fingerprint.

Intensity Function Mathematics
Time-varying Parameters Mathematics
Time series Mathematics
Knot Mathematics
Inhomogeneous Poisson Process Mathematics
Conditional Quantiles Mathematics
Recursive Estimation Mathematics
Non-homogeneous Poisson Process Mathematics

Network Recent external collaboration on country level. Dive into details by clicking on the dots.

Research Output 2006 2020

Dimension estimation using random connection models

Serra, P. & Mandjes, M. R. H., Nov 2017, In : Journal of Machine Learning Research. 18, 35 p., 138.

Research output: Contribution to journalArticleAcademicpeer-review

Estimator
Model
Poisson Model
Dimensionality Reduction
Adjacency Matrix
3 Citations (Scopus)
76 Downloads (Pure)

Adaptive empirical Bayesian smoothing splines

Serra, P. & Krivobokova, T., 2016, In : Bayesian Analysis. 12, 1, p. 219-238 19 p.

Research output: Contribution to journalArticleAcademicpeer-review

Open Access
File
Smoothing Splines
Smoothing Parameter
Empirical Bayes Method
Marginal Likelihood
Regression Function

Estimation of local degree distributions via local weighted averaging and Monte Carlo cross-validation

Serra, P. & Mandjes, M., Apr 2020, In : Computational Statistics and Data Analysis. 144, 21 p., 106886.

Research output: Contribution to journalArticleAcademicpeer-review

Heterogeneous networks
Degree Distribution
Cross-validation
Averaging
Topology
42 Downloads (Pure)
Open Access
File
Replacement Policy
Railroad tracks
Optimal Policy
Railroad
Replacement
1 Citation (Scopus)

Recursive tracking algorithm for a predictable time-varying parameter of a time series

de Andrade Serra, P. J. & Belitser, E. N., 2015, In : Mathematical Methods of Statistics. 24, 4, p. 243-265 22 p.

Research output: Contribution to journalArticleAcademicpeer-review

Time-varying Parameters
Time series
Stochastic Approximation
Stochastic Algorithms
Recursive Algorithm

Courses

Mathematical modelling

1/09/13 → …

Course

Probability and statistics

1/09/13 → …

Course

Programming and modelling

1/09/12 → …

Course

Student theses

Analysis of a sold candy time series with the envelope based approach and Box-Jenkins models

Author: van den Berkmortel, M. P., 7 Aug 2018

Supervisor: De Andrade Serra, P. (Supervisor 1) & Rijpkema, J. (Supervisor 2)

Student thesis: Bachelor

File

Estimation of the global maximiser of noisy functions

Author: Omar, F. M., 17 May 2019

Supervisor: De Andrade Serra, P. (Supervisor 1)

Student thesis: Bachelor

File

Forensic statistics: the decision making process in court from a mathematical point of view

Author: Pasch, E., 12 Jul 2019

Supervisor: De Andrade Serra, P. (Supervisor 1) & Kapodistria, S. (Supervisor 2)

Student thesis: Bachelor

File

Regression models on bivariate count data in an exceptional model mining context

Author: Raaijmakers, B. A., 31 Aug 2017

Supervisor: De Andrade Serra, P. (Supervisor 1) & Duivesteijn, W. (Supervisor 2)

Student thesis: Bachelor

File

Review of discount rates for the valuation of pension liabilities

Author: Theeuwes, I., 28 May 2018

Supervisor: De Andrade Serra, P. (Supervisor 1) & de Bock, H. (External person) (External coach)

Student thesis: Master