-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathreservoirs2.html
345 lines (300 loc) · 10.9 KB
/
reservoirs2.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
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
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
<!DOCTYPE html>
<html lang="en">
<head>
<meta charset="UTF-8">
<meta http-equiv="X-UA-Compatible" content="IE=edge">
<meta name="viewport" content="width=device-width, initial-scale=1.0">
<title>Reservoir Computing</title>
<style>
body {
font-family: Arial, sans-serif;
background-color: #f0f0f0;
margin: 0;
padding: 0;
}
.header {
text-align: center;
padding: 20px;
}
.header img {
width: 200px;
height: 200px;
border-radius: 50%;
}
.header h1 {
margin-top: 10px;
margin-bottom: 0;
}
.header p {
margin-top: 0;
}
.nav {
text-align: center;
background-color: #333;
padding: 10px;
}
.nav a {
color: #fff;
text-decoration: none;
padding: 0 10px;
}
.nav a:hover {
text-decoration: underline;
}
.content {
padding: 20px;
}
.content {
padding: 20px;
background-color: #ffffff; /* this is the color white */
}
body {
font-family: Arial, sans-serif;
background-color: #f0f0f0;
margin: 0;
padding: 0;
}
.section {
display: flex;
align-items: center;
justify-content: space-around;
margin: 30px auto;
padding: 40px;
background-color: #fff;
border-radius: 5px;
box-shadow: 0 0 15px rgba(0,0,0,0.1);
width: 90%;
max-width: 1200px;
}
.section .img1 {
width: 100px;
height: 150px;
}
.section .img2 {
width: 450px;
height: 200px;
}
.section .img3 {
width: 500px;
height: 300px;
}
.section .img4 {
width: 100px;
height: 100px;
}
.section div {
margin-left: 40px;
width: 70%;
}
.section:nth-child(odd) div {
margin-right: 40px;
margin-left: 0;
}
.section:nth-child(odd) {
flex-direction: row-reverse;
}
.intro {
display: flex;
/*align-items: center;*/
/*justify-content: center;*/
margin: 30px auto;
padding: 30px;
background-color: #333;
color: #fff;
border-radius: 5px;
width: 90%;
max-width: 1200px;
/*text-align: center;*/
}
.intro a {
color: #FFD700; /* Change the color of links */
text-decoration: none; /* Remove the underline */
padding: 0px; /* Add some space around the links */
}
.intro a:hover {
color: #fff; /* Change the color when hovered */
text-decoration: underline; /* Underline when hovered */
}
.intro a:first-child {
padding-left: 0;
padding-right: 0;
}
/* CSS for Screen Width up to 600px (Mobile) */
@media only screen and (max-width: 600px) {
.header img {
width: 100px;
height: 100px;
}
.section {
flex-direction: column;
align-items: center;
padding: 20px;
width: 100%;
}
.section .img1, .section .img2, .section .img3, .section .img4 {
width: 100px;
height: auto;
}
.section div {
margin-left: 0;
margin-right: 0;
}
.intro {
flex-direction: column;
align-items: center;
padding: 20px;
width: 100%;
}
}
/* CSS for Screen Width up to 600px (Mobile) */
@media only screen and (max-width: 600px) {
.header img {
width: 100px;
height: 100px;
}
.section {
flex-direction: column;
align-items: center;
padding: 20px;
width: 100%;
}
.section .img1, .section .img2, .section .img3 {
width: 100px;
height: auto;
}
.section div {
margin-left: 0;
margin-right: 0;
}
.intro {
flex-direction: column;
align-items: center;
padding: 20px;
width: 100%;
}
}
.section .img1, .section .img2, .section .img3, .section .img4 {
width: 40%;
height: auto;
}
h1, h2, h4, h5, h6 {
text-align: center;
}
.image-background .esnimg {
background-color: #000;
display: flex;
justify-content: center;
align-items: center;
padding: 0px;
margin: auto;
}
.image-background .esnimg {
max-width: 50%;
height: auto;
}
/* Add styles for the .content paragraph */
.content p {
max-width: 800px;
margin: auto;
padding: 15px;
background-color: rgba(0,0,0,5%);
border-radius: 3px;
}
.content article {
max-width: 900px;
margin: auto;
}
/* Add this code */
.content {
max-width: 1000px;
margin: 0 auto;
}
.content p {
background-color: #f2f2f2; /* Pale gray */
padding: 15px;
border-radius: 5px;
}
.content .paragraph {
background-color: #f2f2f2; /* Pale gray */
padding: 15px;
border-radius: 5px;
margin-bottom: 15px; /* gives some space between paragraphs */
max-width: 800px;
margin-left: auto;
margin-right: auto;
}
.content .paragraph p {
margin: 0;
}
</style>
<link rel="stylesheet" href="style.css">
</head>
<body>
<div class="header">
<img src="portrait.jpg" alt="Your headshot">
<h1>Anthony Polloreno, Ph.D.</h1>
<h2>Research Engineer</h2>
</div>
<div class="nav">
<a href="index.html">Home</a> |
<a href="research.html">Research</a> |
<a href="engineering.html"> Engineering </a> |
<a href="https://twitter.com/ampolloreno">Twitter</a> |
<a href="https://www.linkedin.com/in/ampolloreno/">LinkedIn</a> |
<a href="https://github.com/ampolloreno">GitHub</a> |
<a href="papers.html">Papers</a> |
<a href="resume.pdf">Resume</a>|
<a href="reservoirs.html">Reservoir Computing I</a> |
<a href="reservoirs2.html">Reservoir Computing II</a>|
<a href="reservoirs3.html">Reservoir Computing III</a>
</div>
<div class="content" id="blogPost">
<article>
<h1>The Impact of Noise on Recurrent Neural Networks II</h1>
<div class="paragraph">
<p> In this section, we are going to consider the simulation of the echo state networks discussed in the <a href="reservoirs.html">last post</a>.
This is an oddly constrained problem, and there are actually a few design decisions. Dealing with the variable
training length is annoying. I messed around with this for a while (we will explore this in the appendix),
but discovered that a much better thing to do is simulate an ensemble of reservoirs. In principle, each size
(and each reservoir in each ensemble for each size) should have a different training length, so we don't over- or
under-fit. We still see an improvement with reservoir size if we ignore this, so we do for now. Simulating an
ensemble now gives us a clear dimension to parallelize over. To do this, we now need to chunk the time data,
because loading the entire sequence onto the GPU takes up space that can otherwise be used for simulation and
storing neural activations.
</p>
<p>
To remind ourselves of our final goal, we are interested in simulating echo state networks, and considering their performance
when we also have the output product signals available to us. The final twist is that we are interested in understanding
how noise impacts the performance of these networks. Noise is abundant, and normally protected against in digital logic
by error correcting codes, however if we are interested in building hardware that runs native machine learning algorithms,
there are several reasons why noise may become relevant. First, if the bits per float become small enough, we will see
rounding error - especially in recurrent settings where the errors have a chance to accumulate. If the hardware itself
is physical, or thermodynamic in any sense, thermal noise can cause fluctuations around the expected behavior of the circuits.
If the network is observing data from sources such as sensors, these sensors will often have noise from picking up
undesired signals. In this sense, this is an interesting problem for any real world setting that involves randomness.
</p>
<p>
While this is not the focus of this notebook, we note in passing that this is often the setting in reinforcement learning
contexts in settings such as the REINFORCE algorithm (Williams, Ronald J. "Simple statistical gradient-following algorithms for
connectionist reinforcement learning." Machine learning 8 (1992): 229-256.) where the data being gathered about the world is used
to evaluate a gradient in expectation, which is also a continuous, rather than discrete, value. The REINFORCE algorithm
specifically is a policy gradient method used in reinforcement learning for training policy networks. Given a parameterized
policy, the algorithm approximates the gradient by sampling from the policy policy and using the sampled returns for the
estimation of the gradient. This form of sampling-based estimation allows REINFORCE to update policy parameters in the
direction that, on average, increases the probability of actions that lead to higher returns. That being said, any estimate
of the gradient is imperfect. The imperfections come from finite sample effects in the case that the environment is
stochastic, and come from systematic errors in the case that the policy is a poor representation of the optimal policy.
In this notebook we take a much simpler approach, and simply add Gaussian noise to each value output by the echo state network.
</p>
</div>
<iframe src="Webpost2.html" style="height:10400px; width:100%; border:none;" ></iframe>
<div class="paragraph">
<p>Check out the next notebook <a href="http://ampolloreno.com/reservoirs3.html">here</a>!</p>
<section style="text-align: center;">
<h3>Acknowledgements</h3>
<p>A special thanks to <a href="https://ohaithe.re/">Alex Meiburg</a>, <a href="https://andremelo.org/">André Melo</a> and <a href="https://chromotopy.org/">Eric Peterson</a> for feedback on this post!</p>
</section>
</article>
</div>
</body>
</html>