Menegatto, Aurora (A.A. 2024/2025) Anticipating epileptic seizures through early warning signals: data-driven methods for detecting pre-ictal changes and tipping points in EEG. Tesi di Laurea in Data science in action, Luiss Guido Carli, relatore Alessio Martino, pp. 121. [Master's Degree Thesis]

Preview

PDF (Full text) Download (5MB) | Preview

Abstract/Index

Context description and problem statement. Literature review and dataset description. Tipping points and early warning signals in biomedicine. Datasets and data retrieval. Algorithms implemented in this thesis. Network-based critical transitions across neurological disorders. First methodology: dynamical network modelling via recurrence quantification analysis. Signal preprocessing. Recurrence quantification analysis of dynamic networks. Second methodology: time-frequency modeling via discrete wavelet transforms. Overview of the pipeline. Signal preprocessing. Sliding window segmentation and temporal labelling. Discrete wavelets transform feature extraction. Dataset-specific adaptations: CHB-MIT vs Siena. Feature scaling and class imbalance handling. Hybrid integration of spectral and dynamical features. Context-aware wavelet modeling via RQA. Modeling, experimental results, evaluation and discussion. Feature configurations summary. Modeling and evaluation strategy. Results obtained.

References

Bibliografia: pp. 112-115.

Thesis Type:	Master's Degree Thesis
Institution:	Luiss Guido Carli
Degree Program:	Master's Degree Programs > Master's Degree Program in Data Science e Management (LM-91)
Chair:	Data science in action
Thesis Supervisor:	Martino, Alessio
Thesis Co-Supervisor:	Spagnoletti, Paolo
Academic Year:	2024/2025
Session:	Extraordinary
Deposited by:	Alessandro Perfetti
Date Deposited:	01 Jul 2026 10:31
Last Modified:	01 Jul 2026 10:31
URI:	https://tesi.luiss.it/id/eprint/46268

Downloads

Repository Staff Only

View Item