Advanced Pulse Processing and Crosstalk Mitigation in X-ray Microcalorimeters
(2024)Department of Automatic Control
- Abstract
- This thesis presents an approach to improving pulse processing in singlephoton spectroscopy, focusing on crosstalk detection and correction in x-ray microcalorimeters. The work integrates advanced mathematical models, including Jacobian and Hessian computations, Hamiltonian Neural Networks (HNNs), and the application of negative probabilities to optimize Shannon entropy. These models are embedded within an AI pipeline that leverages PyTorch running on high-performance hardware, comprising an AMD Ryzen Threadripper PRO CPU and a Radeon RX 7900 XTX GPU [AMD, 2023b].
The primary objective of this research is to develop a system capable of identifying and mitigating crosstalk, which is a significant source of noise in photon energy... (More) - This thesis presents an approach to improving pulse processing in singlephoton spectroscopy, focusing on crosstalk detection and correction in x-ray microcalorimeters. The work integrates advanced mathematical models, including Jacobian and Hessian computations, Hamiltonian Neural Networks (HNNs), and the application of negative probabilities to optimize Shannon entropy. These models are embedded within an AI pipeline that leverages PyTorch running on high-performance hardware, comprising an AMD Ryzen Threadripper PRO CPU and a Radeon RX 7900 XTX GPU [AMD, 2023b].
The primary objective of this research is to develop a system capable of identifying and mitigating crosstalk, which is a significant source of noise in photon energy measurements [Hansen and Chervenak, 2014]. Additionally, this work explores the challenges and benefits of using open-source software and ROCm for GPU-accelerated computing in high-performance environments [AMD, 2019].
The thesis contributes to the field by demonstrating a scalable, flexible pipeline that can be adapted for other spectroscopy applications, offering enhanced performance and precision in data analysis. The proposed system not only addresses current limitations in crosstalk correction but also sets the stage for future advancements in data processing and quantum-like system modeling. (Less)
Please use this url to cite or link to this publication:
http://lup.lub.lu.se/student-papers/record/9176325
- author
- Shahin, Abdullah
- supervisor
-
- Jens Uhlig LU
- Karl-Erik Årzén LU
- organization
- year
- 2024
- type
- H3 - Professional qualifications (4 Years - )
- subject
- report number
- TFRT-6259
- other publication id
- 0280-5316
- language
- English
- id
- 9176325
- date added to LUP
- 2024-10-08 15:08:57
- date last changed
- 2024-10-08 15:08:57
@misc{9176325,
abstract = {{This thesis presents an approach to improving pulse processing in singlephoton spectroscopy, focusing on crosstalk detection and correction in x-ray microcalorimeters. The work integrates advanced mathematical models, including Jacobian and Hessian computations, Hamiltonian Neural Networks (HNNs), and the application of negative probabilities to optimize Shannon entropy. These models are embedded within an AI pipeline that leverages PyTorch running on high-performance hardware, comprising an AMD Ryzen Threadripper PRO CPU and a Radeon RX 7900 XTX GPU [AMD, 2023b].
The primary objective of this research is to develop a system capable of identifying and mitigating crosstalk, which is a significant source of noise in photon energy measurements [Hansen and Chervenak, 2014]. Additionally, this work explores the challenges and benefits of using open-source software and ROCm for GPU-accelerated computing in high-performance environments [AMD, 2019].
The thesis contributes to the field by demonstrating a scalable, flexible pipeline that can be adapted for other spectroscopy applications, offering enhanced performance and precision in data analysis. The proposed system not only addresses current limitations in crosstalk correction but also sets the stage for future advancements in data processing and quantum-like system modeling.}},
author = {{Shahin, Abdullah}},
language = {{eng}},
note = {{Student Paper}},
title = {{Advanced Pulse Processing and Crosstalk Mitigation in X-ray Microcalorimeters}},
year = {{2024}},
}