Handbook on Big Data and Machine Learning in the Physical Sciences

Kalinin, Sergei V.; Foster, Ian; Kalidindi, Surya R.; Lookman, Turab; Dam, Kerstin Kleese van; Yager, Kevin G.; Campbell, Stuart I.; Farnsworth, Richard L.; Dam, Maartje van

doi:10.1142/11389-vol2

Cited by 1 publication

(2 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In contrast, MAXI J1535 −571 observations are of varying composition, imbalanced in fa v or of QPO absence. Second, GRS 1915 + 105 has around two times more total observations, and around six times more observations with QPOs than MAXI J1535 −571; in most cases training models on more data leads to corresponding increases in accuracy (Brefeld et al 2020 ;Kalinin & Foster 2020 ). Ho we ver, this assumption may not hold in instances like this, where models are being tested on different objects, as there may exist fundamentally stronger/more pronounced associations between spectral and QPO in one of the systems.…”

Section: Regressionmentioning

confidence: 95%

“…Since we only regress for the fundamental in the GRS 1915 + 105 PDS, its output matrix takes the form OUT = m × 3. Prior to reformatting the data in this manner, we applied a columnar min-max standardization to the XSPEC , and hardness input features, as well as the QPO Lorentzian output features, which linearly transformed each distribution into a [max( x ), min( x )] = [0.1, 1] range (as opposed to the traditional [0 − 1] range given our decision to denote QPO non-detections with zero values) while preserving their shapes, according to equation ( 4 ; Kandanaarachchi et al 2019 ).…”

Section: Feature Engineeringmentioning

confidence: 99%

See 1 more Smart Citation

QPOML: a machine learning approach to detect and characterize quasi-periodic oscillations in X-ray binaries

Kiker

Steiner

Garraffo

et al. 2023

Monthly Notices of the Royal Astronomical Society

View full text Add to dashboard Cite

Astronomy is presently experiencing profound growth in the deployment of machine learning to explore large datasets. However, transient quasi-periodic oscillations (QPOs) which appear in power density spectra of many X-ray binary system observations are an intriguing phenomena heretofore not explored with machine learning. In light of this, we propose and experiment with novel methodologies for predicting the presence and properties of QPOs to make the first ever detections and characterizations of QPOs with machine learning models. We base our findings on raw energy spectra and processed features derived from energy spectra using an abundance of data from the NICER and Rossi X-ray Timing Explorer space telescope archives for two black hole low mass X-ray binary sources, GRS 1915+105 and MAXI J1535-571. We advance these non-traditional methods as a foundation for using machine learning to discover global inter-object generalizations between–and provide unique insights about–energy and timing phenomena to assist with the ongoing challenge of unambiguously understanding the nature and origin of QPOs. Additionally, we have developed a publicly available Python machine learning library, QPOML, to enable further Machine Learning aided investigations into QPOs.

show abstract

Section: Regressionmentioning

confidence: 95%

Section: Feature Engineeringmentioning

confidence: 99%