Di Cesare, Martina (2026) All sky searches for isolated and binary neutron star system: study of improved procedures, based on classical and Machine Learning techniques, for present and future gravitational wave detectors. [Tesi di dottorato]
This is the latest version of this item.
|
Documento PDF
DICESARE_MARTINA_XXXVIII.pdf Visibile a [TBR] Repository staff only Download (25MB) |
|
|
Documento PDF
DICESARE_MARTINA_XXXVIII.pdf Visibile a [TBR] Repository staff only Download (25MB) |
| Item Type: | Tesi di dottorato |
|---|---|
| Resource language: | English |
| Title: | All sky searches for isolated and binary neutron star system: study of improved procedures, based on classical and Machine Learning techniques, for present and future gravitational wave detectors |
| Creators: | Creators Email Di Cesare, Martina martina.dicesare@unina.it |
| Date: | 2026 |
| Number of Pages: | 159 |
| Institution: | Università degli Studi di Napoli Federico II |
| Department: | Fisica |
| Dottorato: | Fisica |
| Ciclo di dottorato: | 38 |
| Coordinatore del Corso di dottorato: | nome email Canale, Vincenzo UNSPECIFIED |
| Tutor: | nome email De Rosa, Rosario UNSPECIFIED Astone, Pia UNSPECIFIED Serra, Marco UNSPECIFIED |
| Date: | 2026 |
| Number of Pages: | 159 |
| Keywords: | Gravitational waves, Neutron stars, Machine Learning |
| Settori scientifico-disciplinari del MIUR: | Area 02 - Scienze fisiche > FIS/05 - Astronomia e astrofisica |
| Additional information: | Dottoranda appartenente al ciclo XVIII. |
| Date Deposited: | 17 Feb 2026 07:24 |
| Last Modified: | 12 Aug 2026 05:37 |
| URI: | https://www.fedoa.unina.it/id/eprint/16183 |
Available Versions of this Item
-
MATERIALI ORGANICI E/O METALLORGANICI PER L’ELETTRONICA E PER L’OPTOELETTRONICA. (deposited 03 Aug 2010 14:31)
- All sky searches for isolated and binary neutron star system: study of improved procedures, based on classical and Machine Learning techniques, for present and future gravitational wave detectors. (deposited 17 Feb 2026 07:24) [Currently Displayed]
Collection description
Gravitational waves (GWs), predicted by General Relativity and first directly detected in 2015, have opened a new observational window on compact objects such as black holes (BHs) and neutron stars (NSs). Among the various classes of GW sources, continuous waves (CWs) - long-lived, nearly monochromatic signals likely emitted by rotating NSs - remain undetected, motivating increasingly sensitive searches with both current and future interferometric detectors. This thesis investigates two complementary directions aimed at improving the detection prospects of CWs in all-sky searches: (i) assessing the performance and limitations of the Frequency Hough (FH) pipeline, which is part of a hierarchical procedure designed to look for unknown isolated NSs, when signals originate from NSs in binary systems, and (ii) developing and testing a Machine-Learning-based alternative to the FH candidates follow-up (FU) stage. (i) For the binary scenario, we quantify the degradation produced by the binary Rømer delay - an additional Doppler effect due to the NS being in a binary system. By exploring various orbital periods, semi-major axes, and sky positions, we show that parameter recovery worsens predictably with stronger binary modulation, though standard coincidence thresholds between detectors remain effective. These results delineate the operational limits of FH for binary CW searches and provide guidance for future observing runs and third-generation detectors, where many known binaries are expected to emit detectable CWs. (ii) Using real O3 coincidence candidates from the two LIGO detectors and software injections at different strain levels, we obtain robust performance, with only a small fraction of errors in signal characterization, and further improvement after cross-detector coincidence. Application to hardware-injection regions reproduces the behaviour of the FH pipeline. Overall, this thesis demonstrates that the FH pipeline, designed to search for isolated NS, can also be used to indicate the presence of a binary CW signals, within a predictable and obviously limited parameter space and that the ML-based strategy offers a feasible and competitive alternative for candidate FU.
Downloads
Downloads per month over past year
Actions (login required)
![]() |
View Item |


