-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathsymbolicRegression.html
78 lines (69 loc) · 3.8 KB
/
symbolicRegression.html
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
<!DOCTYPE html>
<html lang="en">
<head>
<title>Arghya Dutta</title>
<meta charset="utf-8">
<meta name="viewport" content="width=device-width, initial-scale=1, shrink-to-fit=no">
<link rel="apple-touch-icon" sizes="180x180" href="../assets/favicon_io/apple-touch-icon.png">
<link rel="icon" type="image/png" sizes="32x32" href="../assets/favicon_io/favicon-32x32.png">
<link rel="icon" type="image/png" sizes="16x16" href="../assets/favicon_io/favicon-16x16.png">
<link rel="manifest" href="../assets/favicon_io/site.webmanifest">
<link rel="stylesheet" href="../assets/style.css">
<script id="MathJax-script" async src="https://cdn.jsdelivr.net/npm/mathjax@3/es5/tex-mml-chtml.js"> </script>
<script>window.MathJax = { tex: { inlineMath: [['$', '$'], ['\\(', '\\)']] } };</script>
</head>
<em><a href="../notebooks.html">Notebooks</a></em>
<h1>Symbolic Regression</h1>
<p>
A method of systematically generating algebraic equations from data, with potential applications in discovering equations and even laws. The result, so far, has been quite promising, but no "laws" have been discovered, as yet.
</p>
<h2>Timeline of papers</h2>
<dl>
<dt>
Genetic algorithm paper by Koza (1994)
</dt>
<dt>
Schmidt and Lipson (2009). Note this paper has been criticized for not citing the literature (Crutchfield et al. 1987 and 1998).
</dt>
<dt>
SISSO (Luca's group) and subsequent applications in material science.
</dt>
<dt>
AI Feynmann from Tegmark.
</dt>
</dl>
<h2>Recommended</h2>
<dl>
<dt>
Crutchfield, J. P. (1987). <em>Equations of Motion from a Data Series</em>. Complex Systems, 1, 417–452.
</dt>
<dt>
Crutchfield, J. P., & Young, K. (1989). <em>Inferring statistical complexity</em>. <a href="https://doi.org/10.1103/PhysRevLett.63.105">Physical Review Letters, 63(2), 105–108</a>.
</dt>
<dt>
Koza, J. R. (1994). <em>Genetic programming as a means for programming computers by natural selection</em>. <a href="https://doi.org/10.1007/BF00175355">Statistics and Computing, 4(2), 87–112</a>.
</dt>
<dt>
Schmidt, M., & Lipson, H. (2009). <em>Distilling Free-Form Natural Laws from Experimental Data</em>. <a href="https://doi.org/10.1126/science.1165893">Science, 324(5923), 81–85</a>.
</dt>
<dt>
Ouyang, R., Curtarolo, S., Ahmetcik, E., Scheffler, M., & Ghiringhelli, L. M. (2018). <em>SISSO: A compressed-sensing method for identifying the best low-dimensional descriptor in an immensity of offered candidates</em>. <a href="https://doi.org/10.1103/PhysRevMaterials.2.083802">Physical Review Materials, 2(8), 083802</a>.
</dt>
<dt>
Ouyang, R., Ahmetcik, E., Carbogno, C., Scheffler, M., & Ghiringhelli, L. M. (2019). <em>Simultaneous learning of several materials properties from incomplete databases with multi-task SISSO</em>. <a href="https://doi.org/10.1088/2515-7639/ab077b">Journal of Physics: Materials, 2(2), 024002</a>.
</dt>
<dt>
Udrescu, S.-M., & Tegmark, M. (2020). <em>AI Feynman: A Physics-Inspired Method for Symbolic Regression</em> <a href="http://arxiv.org/abs/1905.11481">arXiv:1905.11481</a>.
</dt>
<dt>
Dutta, A., Vreeken, J., Ghiringhelli, L. M., & Bereau, T. (2021). Data-driven equation for drug–membrane permeability across drugs and membranes. The Journal of Chemical Physics, 154(24), 244114. https://doi.org/10.1063/5.0053931
</dt>
<dd>
Yes, that's me. :P
</dd>
<dt>
Purcell, T. A. R., Scheffler, M., Carbogno, C., & Ghiringhelli, L. M. (2022). <em>SISSO++: A C++ Implementation of the Sure-Independence Screening and Sparsifying Operator Approach</em>. <a href="https://doi.org/10.21105/joss.03960">Journal of Open Source Software, 7(71), 3960</a>.
</dt>
</dl>
</body>
</html>