-
Notifications
You must be signed in to change notification settings - Fork 0
/
references.bib
193 lines (187 loc) · 18.5 KB
/
references.bib
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
@article{HighDimLearning,
abstract = {The notion of interpolation and extrapolation is fundamental in various fields from deep learning to function approximation. Interpolation occurs for a sample x whenever this sample falls inside or on the boundary of the given dataset's convex hull. Extrapolation occurs when x falls outside of that convex hull. One fundamental (mis)conception is that state-of-the-art algorithms work so well because of their ability to correctly interpolate training data. A second (mis)conception is that interpolation happens throughout tasks and datasets, in fact, many intuitions and theories rely on that assumption. We empirically and theoretically argue against those two points and demonstrate that on any high-dimensional (>100) dataset, interpolation almost surely never happens. Those results challenge the validity of our current interpolation/extrapolation definition as an indicator of generalization performances.},
author = {Randall Balestriero and Jérôme Pesenti and Yann Lecun},
title = {Learning in High Dimension Always Amounts to Extrapolation},
}
@article{Tardini2023,
abstract = {The recently developed integrated model based on engineering parameters (IMEP) (Luda et al 2020 Nucl. Fusion 61 126048; Luda et al 2021 Nucl. Fusion 60 036023), so far validated on ASDEX Upgrade, has been tested on a database of 3 Alcator C-Mod and 55 JET-ILW ELMy (type I) H-mode stationary phases. The empirical pedestal transport model included in IMEP, consisting now of imposing a fixed value of R < ∇T e > /T e,top = −82.5, allows an accurate prediction of the pedestal top temperature (when the pedestal top density is fixed to the experimental measurements) across these three machines with different sizes, when the pedestal is peeling-ballooning (PB) limited. Cases far from the ideal PB boundary, corresponding to high edge Spitzer resistivity, are instead strongly overpredicted by IMEP. A comparison between the predictions of Europed and IMEP for a subset of JET-ILW cases shows that IMEP can more accurately reproduce the experimental pedestal width. This allows IMEP to better capture profile effects on the pedestal stability, and therefore to correctly describe the negative effect of fueling on the pedestal pressure for PB limited cases. A strong correlation between the separatrix density and the fueling rate has been identified for a subset of JET-ILW cases, when taking into account different divertor configurations. Overall, these promising results encourage further developments of integrated models to obtain reliable predictions of pedestal and global confinement using only engineering parameters for present and future machines. 5 Stroth et al (https://doi.org/10.1088/1741-4326/ac207f) for the ASDEX Upgrade Team 6 Labit et al for the EUROfusion MST1 Team 7 Mailloux et al (https://doi.},
author = {G Tardini and C Angioni and CK Kiefer and T Luda and M G Dunne and E Fable and A Kallenbach and N Bonanomi and P A Schneider and M Siccinio and P Rodriguez-Fernandez and J W Hughes},
doi = {10.1088/1361-6587/acb011},
issue = {8},
keywords = {confinement,integrated modeling,multi-machine validation,pedestal},
title = {Validation of IMEP on Alcator C-Mod and JET-ILW ELMy H-mode plasmas},
volume = {4},
url = {https://doi.org/10.1088/1361-6587/acb011},
year = {2023},
}
@misc{snoek2012practical,
title={Practical Bayesian Optimization of Machine Learning Algorithms},
author={Jasper Snoek and Hugo Larochelle and Ryan P. Adams},
year={2012},
eprint={1206.2944},
archivePrefix={arXiv},
primaryClass={stat.ML}
}
@inproceedings{Rasmussen1995,
author = {Williams, Christopher and Rasmussen, Carl},
booktitle = {Advances in Neural Information Processing Systems},
editor = {D. Touretzky and M.C. Mozer and M. Hasselmo},
pages = {},
publisher = {MIT Press},
title = {Gaussian Processes for Regression},
url = {https://proceedings.neurips.cc/paper_files/paper/1995/file/7cce53cf90577442771720a370c3c723-Paper.pdf},
volume = {8},
year = {1995}
}
@article{Wu2018,
title = {Inverse uncertainty quantification using the modular Bayesian approach based on Gaussian process, Part 1: Theory},
journal = {Nuclear Engineering and Design},
volume = {335},
pages = {339-355},
year = {2018},
issn = {0029-5493},
doi = {https://doi.org/10.1016/j.nucengdes.2018.06.004},
url = {https://www.sciencedirect.com/science/article/pii/S0029549318306423},
author = {Xu Wu and Tomasz Kozlowski and Hadi Meidani and Koroush Shirvan},
keywords = {Inverse uncertainty quantification, Bayesian calibration, Gaussian process, Modular Bayesian, Model discrepancy},
}
@book{RasmussenW06,
author = {Carl Edward Rasmussen and
Christopher K. I. Williams},
title = {Gaussian processes for machine learning},
series = {Adaptive computation and machine learning},
publisher = {{MIT} Press},
year = {2006},
url = {https://www.worldcat.org/oclc/61285753},
isbn = {026218253X},
timestamp = {Fri, 17 Jul 2020 16:12:42 +0200},
biburl = {https://dblp.org/rec/books/lib/RasmussenW06.bib},
bibsource = {dblp computer science bibliography, https://dblp.org}
}
@article{death2021,
abstract = {Bayesian optimisation (BO) uses probabilistic surrogate models-usually Gaussian processes (GPs)-for the optimisation of expensive black-box functions. At each BO iteration, the GP hyperparameters are fit to previously-evaluated data by maximising the marginal likelihood. However, this fails to account for uncertainty in the hy-perparameters themselves, leading to overconfident model predictions. This uncertainty can be accounted for by taking the Bayesian approach of marginalising out the model hyperparameters. We investigate whether a fully-Bayesian treatment of the Gaussian process hyperparameters in BO (FBBO) leads to improved optimisation performance. Since an analytic approach is intractable, we compare FBBO using three approximate inference schemes to the maximum likelihood approach, using the Expected Improvement (EI) and Upper Confidence Bound (UCB) acquisition functions paired with ARD and isotropic Matérn kernels, across 15 well-known benchmark problems for 4 observational noise settings. FBBO using EI with an ARD kernel leads to the best performance in the noise-free setting, with much less difference between combinations of BO components when the noise is increased. FBBO leads to over-exploration with UCB, but is not detrimental with EI. Therefore, we recommend that FBBO using EI with an ARD kernel as the default choice for BO. CCS CONCEPTS • Theory of computation → Gaussian processes; Mathematical optimization; • Mathematics of computing → Bayesian computation.},
author = {George De Ath and Richard M Everson and Jonathan E Fieldsend and Jonathan E 2021 Fieldsend},
doi = {10.1145/3449726.3463164},
isbn = {2105.00894v1},
keywords = {Ap-proximate inference,Bayesian optimisation,Gaussian process,Surrogate modelling},
title = {How Bayesian Should Bayesian Optimisation Be?},
url = {https://doi.org/10.1145/3449726.3463164.},
year = {2021},
}
@phdthesis{Stefanikova2020,
author = {Estera Stefanikova},
isbn = {9789178735211},
title = {Pedestal structure and stability in JET-ILW and comparison with JET-C},
school = {KTH},
year = {2020}
}
@article{Beurskens2014,
abstract = {Type I ELMy H-mode operation in JET with the ITER-like Be/W wall (JET-ILW) generally occurs at lower pedestal pressures compared to those with the full carbon wall (JET-C). The pedestal density is similar but the pedestal temperature where type I ELMs occur is reduced and below to the so-called critical type I-type III transition temperature reported in JET-C experiments. Furthermore, the confinement factor H 98(y,2) in type I ELMy H-mode baseline plasmas is generally lower in JET-ILW compared to JET-C at low power fractions P loss /P thr,08 < 2 (where P loss is (P in − dW /dt), and P thr,08 the L-H power threshold from Martin et al 2008 (J. Phys. Conf. Ser. 123 012033)). Higher power fractions have thus far not been achieved in the baseline plasmas. At P loss /P thr,08 > 2, the confinement in JET-ILW hybrid plasmas is similar to that in JET-C. A reduction in pedestal pressure is the main reason for the reduced confinement in JET-ILW baseline ELMy H-mode plasmas where typically H 98(y,2) = 0.8 is obtained, compared to H 98(y,2) = 1.0 in JET-C. In JET-ILW hybrid plasmas a similarly reduced pedestal pressure is compensated by an increased peaking of the core pressure profile resulting in H 98(y,2) 1.25. The pedestal stability has significantly changed in high triangularity baseline plasmas where the confinement loss is also most apparent. Applying the same stability analysis for JET-C and JET-ILW, the measured pedestal in JET-ILW is stable with respect to the calculated peeling-ballooning stability limit and the ELM collapse time has increased to 2 ms from typically 200 µs in JET-C. This indicates that changes in the pedestal stability may have contributed to the reduced pedestal confinement in JET-ILW plasmas. A comparison of EPED1 pedestal pressure prediction with JET-ILW experimental data in over 500 JET-C and JET-ILW baseline and hybrid plasmas shows a good agreement with 0.8 < (measured p ped)/(predicted p ped,EPED) < 1.2, but that the role of triangularity is generally weaker in the JET-ILW experimental data than in the model predictions.},
author = {M N A Beurskens and L Frassinetti and C Challis and C Giroud and S Saarelma and B Alper and C Angioni and P Bilkova and C Bourdelle and S Brezinsek and P Buratti and G Calabro and T Eich and J Flanagan and E Giovannozzi and M Groth and J Hobirk and E Joffrin and M J Leyland and P Lomas and E De La Luna and M Kempenaars and G Maddison and C Maggi and P Mantica and M Maslov and G Matthews and M.-L Mayoral and R Neu and I Nunes and T Osborne and F Rimini and R Scannell and E R Solano and P B Snyder and I Voitsekhovitch},
doi = {10.1088/0029-5515/54/4/043001},
journal = {| International Atomic Energy Agency Nuclear Fusion Nucl. Fusion},
keywords = {confinement,metal wall,nitrogen,pedestal,radiation,tokamak},
pages = {13},
title = {Global and pedestal confinement in JET with a Be/W metallic wall},
volume = {54},
year = {2014},
}
@article{Saarelma2023,
abstract = {The neutral ionisation model proposed by Groebner et al (2002 Phys. Plasmas 9 2134) to determine the plasma density profile in the H-mode pedestal, is extended to include charge exchange processes in the pedestal stimulated by the ideas of Mahdavi et al (2003 Phys. Plasmas 10 3984). The model is then tested against JET H-mode pedestal data, both in a 'standalone' version using experimental temperature profiles and also by incorporating it in the Europed version of EPED. The model is able to predict the density pedestal over a wide range of conditions with good accuracy. It is also able to predict the experimentally observed isotope effect on the density pedestal that eludes simpler neutral ionization models.},
author = {S Saarelma and JW Connor and P Bilkova and P Bohm and Ar Field and L Frassinetti and R Fridstrom and A Kirk and Jet Contributors},
doi = {10.1088/1741-4326/acc084},
journal = {International Atomic Energy Agency Nuclear Fusion Nucl. Fusion},
keywords = {JET,density,pedestal,prediction},
pages = {52002-52012},
title = {Testing a prediction model for the H-mode density pedestal against JET-ILW pedestals},
volume = {63},
url = {https://doi.org/10.1088/1741-4326/acc084},
year = {2023},
}
@article{Martin_2008,
doi = {10.1088/1742-6596/123/1/012033},
url = {https://dx.doi.org/10.1088/1742-6596/123/1/012033},
year = {2008},
month = {jul},
publisher = {},
volume = {123},
number = {1},
pages = {012033},
author = {Y R Martin and T Takizuka},
title = {Power requirement for accessing the H-mode in ITER},
journal = {Journal of Physics: Conference Series}
}
@inproceedings{XGboost,
author = {Chen, Tianqi and Guestrin, Carlos},
title = {{XGBoost}: A Scalable Tree Boosting System},
booktitle = {Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining},
series = {KDD '16},
year = {2016},
isbn = {978-1-4503-4232-2},
location = {San Francisco, California, USA},
pages = {785--794},
numpages = {10},
url = {http://doi.acm.org/10.1145/2939672.2939785},
doi = {10.1145/2939672.2939785},
acmid = {2939785},
publisher = {ACM},
address = {New York, NY, USA},
keywords = {large-scale machine learning},
}
@article{Ham2021,
abstract = {The physics of the tokamak pedestal is still not fully understood, for example there is no fully predictive model for the pedestal height and width. However, the pedestal is key in determining the fusion power for a given scenario. If we can improve our understanding of reactor relevant pedestals we will improve our confidence in designing potential fusion power plants. Work has been carried out as part of a collaboration on reactor relevant pedestal physics.We report some of the results in detail here and review some of the wider work which will be reported in full elsewhere. First, we attempt to use a gyrokinetic-based calculation to eliminate the pedestal top density as a model input for Europed/EPED pedestal predictions.We assume power balance at the top of the pedestal, that is, the heat flux crossing the separatrix must be equal to the heat source at the top of the pedestal and investigate the consequences of this assumption. Unfortunately, the transport assumptions of the EPED model mean that this method does not discriminate between different pairs of density and temperature profiles for a given pressure profile. Second, we investigate the effects of non flux surface density on the bootstrap current. Third, type I ELMs will not be tolerable for a reactor relevant regime due to the damage that they are expected to cause to plasma facing components. In recent years various methods of running tokamak plasmas without large ELMs have been developed. These include small and no ELM regimes, the use of resonant magnetic perturbations and the use of vertical kicks. We discuss the quiescent H-mode here. Finally we give a summary and directions for future work EURATOM 2021.},
author = {C. J. Ham and A. Bokshi and D. Brunetti and G. B. Ramirez and B. Chapman and J. W. Connor and D. Dickinson and A. R. Field and L. Frassinetti and A. Gillgren and J. P. Graves and T. P. Kiviniemi and S. Leerink7 and B. McMillan and S. Newton and S. Pamela and C. M. Roach and S. Saarelma and J. Simpson and S. F. Smith and E. R. Solano and P. Strand and A. J. Virtanen},
doi = {10.1088/1741-4326/AC12E9},
issn = {17414326},
issue = {9},
journal = {Nuclear Fusion},
keywords = {Mhd,Pedestal,Tokamak},
month = {9},
publisher = {IOP Publishing Ltd},
title = {Towards understanding reactor relevant tokamak pedestals},
volume = {61},
year = {2021},
}
@book{wesson,
author = "Wesson, John",
title = "{Tokamaks; 3th ed.}",
publisher = "Oxford Univ. Press",
address = "Oxford",
series = "International series of monographs on physics",
year = "2011",
url = "https://cds.cern.ch/record/1427009",
}
@misc{ELM_TIMINGS_PREPRINT,
doi = {10.48550/ARXIV.2212.08745},
url = {https://arxiv.org/abs/2212.08745},
author = {O'Shea, Finn H. and Joung, Semin and Smith, David R. and Coffee, Ryan},
keywords = {Plasma Physics (physics.plasm-ph), Data Analysis, Statistics and Probability (physics.data-an), FOS: Physical sciences, FOS: Physical sciences},
title = {Automatic Identification of Edge Localized Modes in the DIII-D Tokamak},
publisher = {arXiv},
year = {2022},
copyright = {Creative Commons Attribution 4.0 International}
}
@article{Frassinetti2021,
abstract = {The EUROfusion JET-ILW pedestal database is described, with emphasis on three main issues. First, the technical aspects are introduced, including a description of the data selection, the datasets, the diagnostics used, the experimental and theoretical methods implemented and the main definitions. Second, the JET-ILW pedestal structure and stability are described. In particular, the work describes the links between the engineering parameters (power, gas and divertor configuration) and the disagreement with the peeling-ballooning (PB) model implemented with ideal magnetohydrodynamics equations. Specifically, the work clarifies why the JET-ILW pedestal tends to be far from the PB boundary at high gas and high power, showing that a universal threshold in power and gas cannot be found but that the relative shift (the distance between the position of the pedestal density and of the pedestal temperature) plays a key role. These links are then used to achieve an empirical explanation of the behavior of the JET-ILW pedestal pressure with gas, power and divertor configuration. Third, the pedestal database is used to revise the scaling law of the pedestal stored energy. The work shows a reasonable agreement with the earlier Cordey scaling in terms of plasma current and triangularity dependence, but highlights some differences in terms of power and isotope mass dependence.},
author = {L. Frassinetti and S. Saarelma and G. Verdoolaege and M. Groth and J. C. Hillesheim and P. Bilkova and P. Bohm and M. Dunne and R. Fridström and E. Giovannozzi and F. Imbeaux and B. Labit and E. De La Luna and C. Maggi and M. Owsiak and R. Scannell},
doi = {10.1088/1741-4326/abb79e},
issn = {17414326},
issue = {1},
journal = {Nuclear Fusion},
title = {Pedestal structure, stability and scalings in JET-ILW: The EUROfusion JET-ILW pedestal database},
volume = {61},
year = {2021},
}
@article{Truong2020,
abstract = {This article presents a selective survey of algorithms for the offline detection of multiple change points in multivariate time series. A general yet structuring methodological strategy is adopted to organize this vast body of work. More precisely, detection algorithms considered in this review are characterized by three elements: a cost function, a search method and a constraint on the number of changes. Each of those elements is described, reviewed and discussed separately. Implementations of the main algorithms described in this article are provided within a Python package called ruptures.},
author = {Charles Truong and Laurent Oudre and Nicolas Vayatis},
doi = {10.1016/J.SIGPRO.2019.107299},
issn = {01651684},
journal = {Signal Processing},
keywords = {Change point detection,Segmentation,Statistical signal processing},
month = {2},
publisher = {Elsevier B.V.},
title = {Selective review of offline change point detection methods},
volume = {167},
year = {2020},
}