Full Text Available
Note: Clicking the button above will open the full text document at the original institutional repository in a new window.
Machine learning approaches towards tuning ALICE TRD simulations
In this work an exploration of the discrepancies existing between real and simulated data pertaining to the ALICE Transition Radiation Detector is carried out as a motivation to tune the necessary parameters in the ALICE Online-Offline simulation software (O2 ). After such exploration a single param...
Saved in:
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Thesis |
| Language: | Eng |
| Published: |
Department of Statistical Sciences
2024
|
| Subjects: | |
| Tags: |
No Tags, Be the first to tag this record!
|
| _version_ | 1867614297229623296 |
|---|---|
| access_status_str | Open Access |
| author | Ramraj, Nikhiel |
| author2 | Dietel, Thomas |
| author_browse | Dietel, Thomas Ramraj, Nikhiel |
| author_facet | Dietel, Thomas Ramraj, Nikhiel |
| author_sort | Ramraj, Nikhiel |
| collection | Thesis |
| description | In this work an exploration of the discrepancies existing between real and simulated data pertaining to the ALICE Transition Radiation Detector is carried out as a motivation to tune the necessary parameters in the ALICE Online-Offline simulation software (O2 ). After such exploration a single parameter namely the Xe gas gain is subjected to modification. A machine learning approach is taken with the use of deep learning discrimination mechanisms namely artificial neural networks and convolutional neural networks to quantify the effect that our tuning has on the improvement of the simulation results and their conformation to the real data. The correspondence of the optimal values suggested by deep learning approaches is investigated with pulse height spectrometry. It is shown that the optimal parameters suggested by our deep learning models through inference of their performance metrics are not clear and in agreement with that suggested by naive pulse height inspections. |
| format | Thesis |
| id | oai:open.uct.ac.za:11427/40362 |
| institution | University of Cape Town (South Africa) |
| language | Eng |
| last_indexed | 2026-06-10T12:49:48.221Z |
| license_str | Not specified — see source repository |
| provenance_str_mv | Harvested via OAI-PMH from UCTD — University of Cape Town Open Access Repository |
| publishDate | 2024 |
| publishDateRange | 2024 |
| publishDateSort | 2024 |
| publisher | Department of Statistical Sciences |
| publisherStr | Department of Statistical Sciences |
| record_format | dspace |
| source_str | UCTD — University of Cape Town Open Access Repository |
| spelling | oai:open.uct.ac.za:11427/40362 Machine learning approaches towards tuning ALICE TRD simulations Ramraj, Nikhiel Dietel, Thomas Statistical Sciences In this work an exploration of the discrepancies existing between real and simulated data pertaining to the ALICE Transition Radiation Detector is carried out as a motivation to tune the necessary parameters in the ALICE Online-Offline simulation software (O2 ). After such exploration a single parameter namely the Xe gas gain is subjected to modification. A machine learning approach is taken with the use of deep learning discrimination mechanisms namely artificial neural networks and convolutional neural networks to quantify the effect that our tuning has on the improvement of the simulation results and their conformation to the real data. The correspondence of the optimal values suggested by deep learning approaches is investigated with pulse height spectrometry. It is shown that the optimal parameters suggested by our deep learning models through inference of their performance metrics are not clear and in agreement with that suggested by naive pulse height inspections. 2024-07-05T12:54:34Z 2024-07-05T12:54:34Z 2024 2024-07-05T12:15:31Z Thesis / Dissertation Masters MSc http://hdl.handle.net/11427/40362 Eng application/pdf Department of Statistical Sciences Faculty of Science |
| spellingShingle | Statistical Sciences Ramraj, Nikhiel Machine learning approaches towards tuning ALICE TRD simulations |
| thesis_degree_str | Master's |
| title | Machine learning approaches towards tuning ALICE TRD simulations |
| title_full | Machine learning approaches towards tuning ALICE TRD simulations |
| title_fullStr | Machine learning approaches towards tuning ALICE TRD simulations |
| title_full_unstemmed | Machine learning approaches towards tuning ALICE TRD simulations |
| title_short | Machine learning approaches towards tuning ALICE TRD simulations |
| title_sort | machine learning approaches towards tuning alice trd simulations |
| topic | Statistical Sciences |
| url | http://hdl.handle.net/11427/40362 |
| work_keys_str_mv | AT ramrajnikhiel machinelearningapproachestowardstuningalicetrdsimulations |