When: 10 March 2025

Where: A1.11Link opens in a new window and LIB 2Link opens in a new window, University of Warwick, Coventry

What: One-day cross-disciplinary workshop on data science

Rationale

Many dynamic phenomena in various environments, from heliophysical plasmas to the stock market and the human brain, appear in the form of predominantly short-lived, highly non-stationary, multi-modal, and nonlinear oscillations. Moreover, observations are often contaminated by noise of natural and instrumental origin. Adequate analysis of such complex dynamic processes requires us to go beyond the assumptions of linearity, stationarity, harmonicity and infinite length of the analysed signal used by traditional Fourier-based techniques. Furthermore, the role of machine learning in time series and image processing in various applications continues to grow. The aim of this one-day cross-disciplinary workshop is to provide a platform for experts from academia and industry to showcase their state-of-the-art in data analysis and forecasting, discuss outstanding problems, and boost the exchange of ideas and collaborations. The organisation of this workshop is supported by the Higher Education Innovation Funding (HEIF), Space Science Impact Fund (SSIF) grant, EPSRC EP/Y037456/1 grant, and Physics Day funding.

Participating Organisations

• a.i. solutionsLink opens in a new window, NASA Kennedy Space Center
• LEONI Wiring Systems UK Ltd.Link opens in a new window
• Centre for Human Brain HealthLink opens in a new window, University of Birmingham

• Astrophysics Research CentreLink opens in a new window, Queen’s University Belfast

• Instituto de Astronomia Teorica y ExperimentalLink opens in a new window, CONICET-UNC

• Signal and Information Processing LabLink opens in a new window, University of Warwick
• School of EngineeringLink opens in a new window, University of Warwick

Programme

09:00-12:00 Morning Session "Transform-based Methods" (A1.11Link opens in a new window, Chair: Valery Nakariakov)

09:00 – 09:25 Shahin Jafarzadeh (Queen’s University Belfast) – Wave analysis tools (25 min)

09:25 – 09:40 Rebecca Meadowcroft (University of Warwick) – Evaluating solar imaging data with time-distance analysis (15 min)

09:40 – 10:05 Tim Duckenfield (Queen’s University Belfast) – Motion Magnification: Revealing hidden dynamics in solar oscillations and beyond (25 min)

10:05 – 10:25 Vladislavs Bezrukovs (Ventspils University of Applied Sciences) – Challenges in correlation and imaging of very long baseline radio astronomical and space data (20 min)

10:25 – 10:50 Rajesh Govindan (University of Warwick) – Assessment of subjective discomfort in semi-supine humans in a vibrational environment (25 min)

10:50 – 11:05 Weijie Gu (Imperial College London) – Least-squares method for best-fitting Fourier power spectra (15 min)

11:05 – 11:20 Anhad Bagga (University of Warwick) – Error estimation in the Fourier domain using Monte Carlo simulations (15 min)

11:20 – 11:45 Andrew Quinn (University of Birmingham) – A different perspective on neuronal oscillations with Empirical Mode Decomposition (25 min)

11:45 – 12:00 Dmitrii Kolotkov (University of Warwick) – SCOPE: Statistical confidence of oscillatory processes with EMD (15 min)

12:00-14:00 Lunch, Networking & Engagement (Physics ConcourseLink opens in a new window)

14:00-16:00 Afternoon Session "ML and Big Data Methods" (LIB 2Link opens in a new window, Chair: Anne-Marie Broomhall)

14:00 – 14:25 Tim Duckenfield (Queen’s University Belfast) – Hands-on tutorial on Motion Magnification (25 min)

14:25 – 14:50 Bhushan Atote (University of Warwick) – Enhancing explainability in AI: prototype-based insights for image classification and generative models (25 min)

14:50 – 15:05 Sergey Belov (University of Warwick) – FCN for classifying stellar light curves (15 min)

15:05 – 15:30 Katie Kosak (a.i. solutions / NASA Kennedy Space Center) – Space weather data in the launch commit criteria (25 min, online)

15:30 – 15:55 Richard Nederlander (Aegis Aerospace / NASA Kennedy Space Center) – Big Data handling techniques for space weather (25 min, online)

Registration FormLink opens in a new window

Book of AbstractsLink opens in a new window

Toolkits discussed at the workshop:

• Wave analysis tool: WaLSAtoolsLink opens in a new window
• DTCWT based motion magnificationLink opens in a new window and tutorialLink opens in a new window
• 2D Inverse Fourier Transform PlaygroundLink opens in a new window
• Empirical Mode Decomposition in PythonLink opens in a new window
• Statistical Confidence of Oscillatory Processes with EMD: SCOPELink opens in a new window
• Fully Convolutional NetworkLink opens in a new window for detecting oscillations in solar and stellar flare lightcurves

Relevant links

UoW Research & Impact ServicesLink opens in a new window

HEIF/UK Research and InnovationLink opens in a new window

STFC/UK Research and InnovationLink opens in a new window

UoW Interactive Campus MapLink opens in a new window

Scientific Organising Committee (SOC)

Dmitrii Kolotkov (PI of SSIF project, Chair), D.Kolotkov.1@warwick.ac.uk

Valery Nakariakov

Anne-Marie Broomhall

Sergey Belov

Weijie Gu

Feedback from participants

"I enjoyed the workshop very much, and would love more opportunities for this kind of interdisciplinary knowledge sharing and collaboration."

3rd-year PhD

"It was an engaging and insightful event that brought together researchers from diverse fields, fostering meaningful interactions and networking. Incorporating hands-on sessions and allowing more time for discussions could further enhance future editions. Overall, a successful and stimulating event—kudos to the organizers!"

Early Career Researcher

"Excellent and well-organised event. I unfortunately could not attend due to teaching commitments, but the professionalism of the meeting and hosting was fantastic. High quality event and brilliant to bring together multi-disciplinary researchers under the same umbrella."

Professor

"Thank you very much for a very well-organised and productive meeting. Such cross-disciplinary discussions on wave studies are essential for collaborative innovation and technique development, and this meeting was a great step towards in this direction."

Research Fellow

"Perfect work!"

Researcher