-
Notifications
You must be signed in to change notification settings - Fork 9
/
topics.txt
107 lines (57 loc) · 2.56 KB
/
topics.txt
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
* few body problem
* energy conservation
* sympletic integrators / Leap frog
* close encounters / softening
* look at Wisdom-Holman -- decomposes orbit into Keplerian part and interaction part. See,
e.g., https://academic.oup.com/mnras/article/452/1/376/1748797
* autodiff
do a very simple autodiff before machine learning
* reaction networks
* stiff ODEs
* stellar structure
* ODEs
* shooting
* self-gravity
* multipole method
* hydro
* subgrid modeling with Burgers equation
* fitting
* orthogonal distance regression: https://docs.scipy.org/doc/scipy/reference/odr.html#module-scipy.odr
* bayesian
see https://bayesiancomputationbook.com/welcome.html
maybe create fake RV data from a system with 4 planets (with some inclination)
and try to recover their orbital periods?
here's a nice fit to a line example: https://arxiv.org/pdf/1008.4686.pdf
Alex Ji: https://github.com/alexji/cda-samplers
https://arxiv.org/abs/1802.06810
Fitting a line: https://github.com/bmorris3/mcmc_interact
* machine learning
instead of MNIST, try Cifar10
sample dataset:
* galaxies: https://zenodo.org/record/4288837#.YYTamy2l1Z0
* GZ2, GZ-DECaLs (Galaxy Zoo):
* https://zenodo.org/record/3565489
* https://zenodo.org/record/4573248
* better is galaxy10 from astroNN:
* https://astronn.readthedocs.io/en/latest/galaxy10.html
* https://towardsdatascience.com/galaxy-multi-image-classification-with-lenet-5-1bc2f66d3cfc
See also: https://en.wikipedia.org/wiki/List_of_datasets_for_machine-learning_research#Astronomy
Also the AstroML sample datasets: https://www.astroml.org/examples/datasets/index.html
hello universe:
https://archive.stsci.edu/hello-universe
Emily Hunt: https://github.com/emilyhunt/machine-learning-tutorial-MWGaia-DN
* FFT
for C++, consider using pocketFFT -- this is C++ header, and is used by NumPy
https://gitlab.mpcdf.mpg.de/mtr/pocketfft/tree/cpp
we can pass in the `.data()` of the underlying vector.
* Fitting an orbit:
See this paper for grid search: https://ui.adsabs.harvard.edu/abs/1989AJ.....98.1014H/abstract
and this: https://ui.adsabs.harvard.edu/abs/1999AJ....117.1023M/abstract
nice example: https://docs.exoplanet.codes/en/v0.4.5/tutorials/astrometric/
* Gaussian processes:
* review: https://arxiv.org/pdf/2209.08940.pdf
* https://github.com/bmorris3/gp_interact
* diffusion:
Do the simple accretion disk diffusion model
dripping handrail model:
https://articles.adsabs.harvard.edu/pdf/1996ApJ...468..617Y