Ronaldas Macas

Data Scientist


I am interested in data science problems. An expert in time-series analysis, signal processing, statistical modelling and machine-learning applications. Previously analysed data from the most sensitive instrument in the world to search for exploding stars.


  • Machine learning
  • Statistical modelling
  • Code development
  • Gravitational waves


  • PhD in Astrophysics, 2016-2020

    Cardiff University

  • MSci in Physics and Astronomy, 2011-2016

    University of Glasgow

Work Experience


Data Scientist

University of Porstmouth

Sep 2023 – Dec 2023 Portsmouth, United Kingdom
Analysing audio, GPS and temperature data from the GB Row Challenge 2023:

  • Identified underwater noise levels around Great Britain using Python, scikit-learn, and spectral analysis
  • Created a machine-learning algorithm to detect dolphins and ship engines in the audio data resulting in 40% more detections
  • Reduced false alarms by 20% by developing an automated tool to recognize when the microplastic pump was turned on while rowing

Postdoctoral Research Fellow

Institute of Cosmology and Gravitation

Sep 2020 – Aug 2023 Portsmouth, United Kingdom
Responsibilities include:

  • Code development in Python, JAX, PyMC, TensorFlow and PyTorch
  • Leading and training a group of ~30 people to investigate gravitational-wave data quality around astrophysical events
  • Closely interacting and working with various research groups within the LIGO-Virgo-KAGRA scientific collaboration to build automated gravitational-wave data quality tools
  • Student supervision
  • Public outreach

PhD student in Gravitational Physics

Cardiff University

Oct 2016 – Aug 2020 Portsmouth, United Kingdom
Responsibilities include:

  • Code development in Python and Matlab
  • Leading and training a group of ~10 people to search for gravitational waves associated with Gamma-ray bursts
  • Presenting the group’s work within the LIGO-Virgo-KAGRA scientific collaboration, at various seminars and international conferences
  • Student teaching
  • Public outreach
  • Men’s Basketball Club president


Fast glitch modelling

Modelling gravitational-wave glitches with autoencoders.


Probabilistic modelling of glitches in gravitational-wave data.

Probabilistic noise estimation

Measuring the amount of non-Gaussian noise in the data.

Sky localisation

The effect of noise in localising gravitational-wave events.

Non-linear noise subtraction

Broadband noise modelling with dense neural networks.

Gamma-ray bursts

Searching for gravitational-waves associated with Gamma-ray bursts.


Producing publicly available galaxy catalogue.


Evaluating the performance of BayesWave.

Selected Talks

Revisiting GW200129 with machine learning noise mitigation: is it (still) precessing?

Binary black hole GW200129 is claimed to be the highest precessing binary ever observed. In fact, the measured orbital precession is …

Gravitational-wave detector data quality

Gravitational-wave transient noise modelling and mitigation

Gravitational-wave transient noise poses a challenge in the upcoming LIGO-Virgo-KAGRA fourth observing (O4) run. In this talk, I will …

Do glitches in gravitational-wave data affect our ability to estimate the correct sky localization?

Gravitational-wave (GW) data contains non-Gaussian noise artifacts called ‘glitches’. These glitches can sometimes overlap …