-
Notifications
You must be signed in to change notification settings - Fork 1
/
escape_perc.diff
281 lines (281 loc) · 16.3 KB
/
escape_perc.diff
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
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
Index: article.bib
===================================================================
--- article.bib (revision 376)
+++ article.bib (working copy)
@@ -365,7 +365,7 @@
robust formulation based on the Student’s t-distribution of the
error. We demonstrate how the corresponding penalty function, together
with the sampling approach, can obtain good results for a large-scale
- seismic inverse problem with 50 % corrupted data.},
+ seismic inverse problem with 50 \% corrupted data.},
keywords = {Inverse problems, Seismic inversion, Stochastic optimization, Robust
estimation, Optimization, FWI},
optdoi = {10.1007/s10107-012-0571-6},
Index: conference.bib
===================================================================
--- conference.bib (revision 376)
+++ conference.bib (working copy)
@@ -774,7 +774,7 @@
sources and/or receivers by exploiting the
multidimensional dependencies in the data. We are
able to recover data volumes amidst extremely high
- subsampling ratios (in some cases, > 75%) using this
+ subsampling ratios (in some cases, > 75\%) using this
approach.},
keywords = {EAGE, structured tensor, 3D data interpolation, riemannian optimization},
url = {https://slim.gatech.edu/Publications/Public/Conferences/EAGE/2013/dasilva2013EAGEhtucktensor/dasilva2013EAGEhtucktensor.pdf},
@@ -980,7 +980,7 @@
demonstrate that in our application, a
preconditioner bound to one processor core and
accessing memory contiguously reduces execution time
- by 7% for matrices having on the order of 108
+ by 7\% for matrices having on the order of 108
non-zeros. For reference we note that our C
implementation is over 80 times faster than the
corresponding code written for a high-level
@@ -1174,8 +1174,8 @@
DR. The efficacy of the methods are demonstrated for a large-scale
seismic inverse problem using the robust Student's t-distribution,
where a useful synthetic velocity model is recovered in the extreme
- scenario of 60% corrupted data. The sampling approach achieves this
- recovery using 20% of the effort required by a direct robust approach.},
+ scenario of 60\% corrupted data. The sampling approach achieves this
+ recovery using 20\% of the effort required by a direct robust approach.},
keywords = {ICASSP},
doi = {10.1109/ICASSP.2012.6289103},
url = {https://slim.gatech.edu/Publications/Public/Conferences/ICASSP/2012/AravkinFriedlanderLeeuwen/AravkinFriedlanderLeeuwen.pdf }
Index: internal.bib
===================================================================
--- internal.bib (revision 376)
+++ internal.bib (working copy)
@@ -774,7 +774,7 @@
sources and/or receivers by exploiting the
multidimensional dependencies in the data. We are
able to recover data volumes amidst extremely high
- subsampling ratios (in some cases, > 75%) using this
+ subsampling ratios (in some cases, > 75\%) using this
approach.},
keywords = {EAGE, structured tensor, 3D data interpolation, riemannian optimization},
url = {https://slim.gatech.edu/Publications/Public/Conferences/EAGE/2013/dasilva2013EAGEhtucktensor/dasilva2013EAGEhtucktensor.pdf},
@@ -980,7 +980,7 @@
demonstrate that in our application, a
preconditioner bound to one processor core and
accessing memory contiguously reduces execution time
- by 7% for matrices having on the order of 108
+ by 7\% for matrices having on the order of 108
non-zeros. For reference we note that our C
implementation is over 80 times faster than the
corresponding code written for a high-level
@@ -1174,8 +1174,8 @@
DR. The efficacy of the methods are demonstrated for a large-scale
seismic inverse problem using the robust Student's t-distribution,
where a useful synthetic velocity model is recovered in the extreme
- scenario of 60% corrupted data. The sampling approach achieves this
- recovery using 20% of the effort required by a direct robust approach.},
+ scenario of 60\% corrupted data. The sampling approach achieves this
+ recovery using 20\% of the effort required by a direct robust approach.},
keywords = {ICASSP},
doi = {10.1109/ICASSP.2012.6289103},
url = {https://slim.gatech.edu/Publications/Public/Conferences/ICASSP/2012/AravkinFriedlanderLeeuwen/AravkinFriedlanderLeeuwen.pdf }
@@ -7511,7 +7511,7 @@
sources and/or receivers by exploiting the
multidimensional dependencies in the data. We are
able to recover data volumes amidst extremely high
- subsampling ratios (in some cases, > 75%) using this
+ subsampling ratios (in some cases, > 75\%) using this
approach and we demonstrate our recovery on a
synthetic 5D data set provided to us by BG.},
keywords = {Presentation, SINBAD, SINBADSPRING2013, SLIM,
@@ -8535,7 +8535,7 @@
velocity model was obtained using standard PSDM
model building including anisotropic reflection
tomography, and contained epsilon values as high as
- 20%. We have also attempted full-elastic inversion
+ 20\%. We have also attempted full-elastic inversion
of these data to recover a shallow isotropic model
of both p and s-wave velocities. The final FWI
velocity model shows a network of shallow
@@ -9376,7 +9376,7 @@
reciprocity. Using our method we were able to
achieve results with a 20.45 dB signal to noise
ratio when reconstructing a marine data set that had
- 50% of its traces decimated. This is a 13.44 dB
+ 50\% of its traces decimated. This is a 13.44 dB
improvement over using the same method run without
taking reciprocity into account.},
keywords = {MSc, thesis},
@@ -9850,7 +9850,7 @@
demonstrate that in our application, a
preconditioner bound to one processor core and
accessing memory contiguously reduces execution time
- by 7% for matrices having on the order of 108
+ by 7\% for matrices having on the order of 108
non-zeros. For reference we note that our C
implementation is over 80 times faster than the
corresponding code written for a high-level
@@ -10712,7 +10712,7 @@
compared to weighted $\ell_1$ minimization. Moreover, the
sufficient recovery conditions of weighted $\ell_p$ are
weaker than those of regular $\ell_p$ minimization if at
- least 50% of the support estimate is accurate. We
+ least 50\% of the support estimate is accurate. We
also review some algorithms which exist to solve the
non-convex $\ell_p$ problem and illustrate our results
with numerical experiments.},
@@ -11243,7 +11243,7 @@
robust formulation based on the Student’s t-distribution of the
error. We demonstrate how the corresponding penalty function, together
with the sampling approach, can obtain good results for a large-scale
- seismic inverse problem with 50 % corrupted data.},
+ seismic inverse problem with 50 \% corrupted data.},
keywords = {Inverse problems, Seismic inversion, Stochastic optimization, Robust
estimation, Optimization, FWI},
optdoi = {10.1007/s10107-012-0571-6},
Index: masterthesis.bib
===================================================================
--- masterthesis.bib (revision 376)
+++ masterthesis.bib (working copy)
@@ -62,7 +62,7 @@
reciprocity. Using our method we were able to
achieve results with a 20.45 dB signal to noise
ratio when reconstructing a marine data set that had
- 50% of its traces decimated. This is a 13.44 dB
+ 50\% of its traces decimated. This is a 13.44 dB
improvement over using the same method run without
taking reciprocity into account.},
keywords = {MSc, thesis},
Index: presentation.bib
===================================================================
--- presentation.bib (revision 376)
+++ presentation.bib (working copy)
@@ -979,7 +979,7 @@
sources and/or receivers by exploiting the
multidimensional dependencies in the data. We are
able to recover data volumes amidst extremely high
- subsampling ratios (in some cases, > 75%) using this
+ subsampling ratios (in some cases, > 75\%) using this
approach and we demonstrate our recovery on a
synthetic 5D data set provided to us by BG.},
keywords = {Presentation, SINBAD, SINBADSPRING2013, SLIM,
@@ -2003,7 +2003,7 @@
velocity model was obtained using standard PSDM
model building including anisotropic reflection
tomography, and contained epsilon values as high as
- 20%. We have also attempted full-elastic inversion
+ 20\%. We have also attempted full-elastic inversion
of these data to recover a shallow isotropic model
of both p and s-wave velocities. The final FWI
velocity model shows a network of shallow
Index: slimbib.bib
===================================================================
--- slimbib.bib (revision 376)
+++ slimbib.bib (working copy)
@@ -774,7 +774,7 @@
sources and/or receivers by exploiting the
multidimensional dependencies in the data. We are
able to recover data volumes amidst extremely high
- subsampling ratios (in some cases, > 75%) using this
+ subsampling ratios (in some cases, > 75\%) using this
approach.},
keywords = {EAGE, structured tensor, 3D data interpolation, riemannian optimization},
url = {https://slim.gatech.edu/Publications/Public/Conferences/EAGE/2013/dasilva2013EAGEhtucktensor/dasilva2013EAGEhtucktensor.pdf},
@@ -980,7 +980,7 @@
demonstrate that in our application, a
preconditioner bound to one processor core and
accessing memory contiguously reduces execution time
- by 7% for matrices having on the order of 108
+ by 7\% for matrices having on the order of 108
non-zeros. For reference we note that our C
implementation is over 80 times faster than the
corresponding code written for a high-level
@@ -1174,8 +1174,8 @@
DR. The efficacy of the methods are demonstrated for a large-scale
seismic inverse problem using the robust Student's t-distribution,
where a useful synthetic velocity model is recovered in the extreme
- scenario of 60% corrupted data. The sampling approach achieves this
- recovery using 20% of the effort required by a direct robust approach.},
+ scenario of 60\% corrupted data. The sampling approach achieves this
+ recovery using 20\% of the effort required by a direct robust approach.},
keywords = {ICASSP},
doi = {10.1109/ICASSP.2012.6289103},
url = {https://slim.gatech.edu/Publications/Public/Conferences/ICASSP/2012/AravkinFriedlanderLeeuwen/AravkinFriedlanderLeeuwen.pdf }
@@ -7511,7 +7511,7 @@
sources and/or receivers by exploiting the
multidimensional dependencies in the data. We are
able to recover data volumes amidst extremely high
- subsampling ratios (in some cases, > 75%) using this
+ subsampling ratios (in some cases, > 75\%) using this
approach and we demonstrate our recovery on a
synthetic 5D data set provided to us by BG.},
keywords = {Presentation, SINBAD, SINBADSPRING2013, SLIM,
@@ -8535,7 +8535,7 @@
velocity model was obtained using standard PSDM
model building including anisotropic reflection
tomography, and contained epsilon values as high as
- 20%. We have also attempted full-elastic inversion
+ 20\%. We have also attempted full-elastic inversion
of these data to recover a shallow isotropic model
of both p and s-wave velocities. The final FWI
velocity model shows a network of shallow
@@ -9376,7 +9376,7 @@
reciprocity. Using our method we were able to
achieve results with a 20.45 dB signal to noise
ratio when reconstructing a marine data set that had
- 50% of its traces decimated. This is a 13.44 dB
+ 50\% of its traces decimated. This is a 13.44 dB
improvement over using the same method run without
taking reciprocity into account.},
keywords = {MSc, thesis},
@@ -9850,7 +9850,7 @@
demonstrate that in our application, a
preconditioner bound to one processor core and
accessing memory contiguously reduces execution time
- by 7% for matrices having on the order of 108
+ by 7\% for matrices having on the order of 108
non-zeros. For reference we note that our C
implementation is over 80 times faster than the
corresponding code written for a high-level
@@ -10712,7 +10712,7 @@
compared to weighted $\ell_1$ minimization. Moreover, the
sufficient recovery conditions of weighted $\ell_p$ are
weaker than those of regular $\ell_p$ minimization if at
- least 50% of the support estimate is accurate. We
+ least 50\% of the support estimate is accurate. We
also review some algorithms which exist to solve the
non-convex $\ell_p$ problem and illustrate our results
with numerical experiments.},
@@ -11243,7 +11243,7 @@
robust formulation based on the Student’s t-distribution of the
error. We demonstrate how the corresponding penalty function, together
with the sampling approach, can obtain good results for a large-scale
- seismic inverse problem with 50 % corrupted data.},
+ seismic inverse problem with 50 \% corrupted data.},
keywords = {Inverse problems, Seismic inversion, Stochastic optimization, Robust
estimation, Optimization, FWI},
optdoi = {10.1007/s10107-012-0571-6},
Index: techreport.bib
===================================================================
--- techreport.bib (revision 376)
+++ techreport.bib (working copy)
@@ -236,7 +236,7 @@
demonstrate that in our application, a
preconditioner bound to one processor core and
accessing memory contiguously reduces execution time
- by 7% for matrices having on the order of 108
+ by 7\% for matrices having on the order of 108
non-zeros. For reference we note that our C
implementation is over 80 times faster than the
corresponding code written for a high-level
Index: unpublished.bib
===================================================================
--- unpublished.bib (revision 376)
+++ unpublished.bib (working copy)
@@ -431,7 +431,7 @@
compared to weighted $\ell_1$ minimization. Moreover, the
sufficient recovery conditions of weighted $\ell_p$ are
weaker than those of regular $\ell_p$ minimization if at
- least 50% of the support estimate is accurate. We
+ least 50\% of the support estimate is accurate. We
also review some algorithms which exist to solve the
non-convex $\ell_p$ problem and illustrate our results
with numerical experiments.},