-
Notifications
You must be signed in to change notification settings - Fork 95
/
Copy pathneuroevolution.html
232 lines (217 loc) · 8.61 KB
/
neuroevolution.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
<!doctype html>
<html>
<head>
<meta charset="utf-8">
<meta name="viewport" content="width=device-width, initial-scale=1.0, maximum-scale=1.0, user-scalable=no">
<title>4.3 Evolution of Programs, Graphs, and Networks</title>
<link rel="stylesheet" href="../reveal/css/reset.css">
<link rel="stylesheet" href="../reveal/css/reveal.css">
<link rel="stylesheet" href="../reveal/css/theme/evo.css">
<!-- <link rel="stylesheet" href="../reveal/css/custom.css"> -->
<!-- Theme used for syntax highlighting of code -->
<link rel="stylesheet" href="../reveal/lib/css/zenburn.css">
<!-- Printing and PDF exports -->
<script>
var link = document.createElement( 'link' );
link.rel = 'stylesheet';
link.type = 'text/css';
link.href = window.location.search.match( /print-pdf/gi ) ? '../reveal/css/print/pdf.css' : '../reveal/css/print/paper.css';
document.getElementsByTagName( 'head' )[0].appendChild( link );
</script>
</head>
<body>
<div class="reveal">
<div class="slides">
<section>
<h1>Evolutionary Computation</h1>
<h3>Evolution of Neural Networks</h3>
<br />
<img src="../imgs/logo.png" width="30%" height="auto">
</section>
<section>
<h2>Motivations for neuroevolution</h2>
<img src="../imgs/intro_neuroevolution.png" width="50%">
<br/>
<br/>
<div class="textbox">
<small>
Miikkulainen, Risto. "Evolution of neural networks." Proceedings of the Genetic and Evolutionary Computation Conference Companion. 2017.
</small>
</div>
</section>
<section>
<h2>Neuron Model</h2>
<img data-src="../imgs/neuron.jpg" width="50%" height="auto">
<p class="textbox">
<small>
C. Stangor and J. Walinga. Introduction to Psychology.<br />[BC open textbook
collection]. Flat World Knowledge, L.L.C., 2015.
</small>
</p>
</section>
<section>
<h2>Artificial Neural Networks</h2>
<img data-src="../imgs/neuron_model.jpeg" width="40%" height="auto">
<br />
<img data-src="../imgs/cnn.png" width="40%" height="auto">
<p class="textbox">
<small>
<code>cs231n.github.io</code>
</small>
</p>
</section>
<section>
<h2>Neuroevolution</h2>
<table class="stretch" border="0" width="70%">
<tr>
<td>
<img data-src="../imgs/cellular_encoding.png" width="100%" height="auto">
<br/>
<small>
Gruau, Frederic, and Darrell Whitley. "Adding learning
to the cellular development of neural networks:
Evolution and the Baldwin effect." Evolutionary
computation 1.3 (1993): 213-233.
</small>
</td>
<td valign="center">
<img data-src="../imgs/barr.png" width="100%" height="auto">
<br/>
<small>
Fleischer, Kurt, and Alan H. Barr. "A simulation testbed
for the study of multicellular development: The multiple
mechanisms of morphogenesis." 1994
</small>
</td>
</tr>
</table>
</section>
<section>
<h2>Neuroevolution</h2>
<table class="stretch" border="0" width="70%">
<tr>
<td>
<img data-src="../imgs/neat.png" width="100%" height="auto">
<br/>
<small>
Stanley, Kenneth O., and Risto Miikkulainen. "Evolving
neural networks through augmenting topologies."
Evolutionary computation 10.2 (2002): 99-127
</small>
</td>
<td>
<img data-src="../imgs/cgpbrain-page1.png" width="50%" height="auto">
<br/>
<small>
Miller, Julian F., Dennis G. Wilson, and Sylvain
Cussat-Blanc. "Evolving Developmental Programs That
Build Neural Networks for Solving Multiple Problems."
Genetic Programming Theory and Practice XVI. Springer,
Cham, 2019. 137-178.
</small>
</td>
</tr>
</table>
</section>
<section>
<h2>Neuroevolution</h2>
<img src="../imgs/ne_basics.png" width="50%">
<br/>
<div class="textbox">
<small>
Miikkulainen, Risto. "Evolution of neural networks." Proceedings of the Genetic and Evolutionary Computation Conference Companion. 2017.
</small>
</div>
</section>
<section>
<h2>Evolution of Neural Structure</h2>
<img src="../imgs/ne_structure.png" width="50%">
<br/>
<div class="textbox">
<small>
Miikkulainen, Risto. "Evolution of neural networks." Proceedings of the Genetic and Evolutionary Computation Conference Companion. 2017.
</small>
</div>
</section>
<section>
<h2>Problems with Neuroevolution</h2>
<img src="../imgs/ne_problems.png" width="40%">
<br/>
<div class="textbox">
<small>
Miikkulainen, Risto. "Evolution of neural networks." Proceedings of the Genetic and Evolutionary Computation Conference Companion. 2017.
</small>
</div>
</section>
<section>
<h2>Direct Encoding</h2>
<img src="../imgs/neat.png" width="30%">
<br/>
<br/>
Neural network structure or weights are directly encoded in genome: NEAT
<br/>
<br/>
<div class="textbox">
<small>Stanley, Kenneth O., and Risto Miikkulainen. "Evolving neural networks through augmenting topologies." Evolutionary computation 10.2 (2002): 99-127
</small>
</div>
</section>
<section>
<h2>Indirect Encoding</h2>
<img src="../imgs/cellular_encoding_2.png" width="40%">
<br/>
Neural network structure or weights are the result of developing the genome: Cellular Encoding, HyperNEAT
<br/>
<div class="textbox">
<small>
Miikkulainen, Risto. "Evolution of neural networks." Proceedings of the Genetic and Evolutionary Computation Conference Companion. 2017, from Gruau, Frederic, and Darrell Whitley. "Adding learning to the cellular development of neural networks: Evolution and the Baldwin effect." Evolutionary computation 1.3 (1993): 213-233.
</small>
</div>
</section>
<section>
<h2>Evolution of Deep Neural Networks</h2>
<img src="../imgs/openai_pong.png" width="30%">
<img src="../imgs/openai_noise.png" width="30%">
<br/>
<br/>
Weight optimization only using Evolutionary Strategies
<br/>
<br/>
<div class="textbox">
<small>Salimans, Tim, et al. "Evolution strategies as a scalable alternative to reinforcement learning." arXiv preprint arXiv:1703.03864 (2017).</small>
</div>
</section>
</div>
<div id="footer-container" style="display:none;">
<div id="footer">
Evolutionary Computation by Dennis G. Wilson, Yuri Lavinas, Paul Templier
<br />
<a href="https://github.com/d9w/evolution/">https://github.com/d9w/evolution/</a>
<br />
<a rel="license" href="http://creativecommons.org/licenses/by-sa/4.0/"><img alt="Creative Commons License" style="border-width:0" src="https://i.creativecommons.org/l/by-sa/4.0/80x15.png" /></a>
</div>
</div>
</div>
<script src="../reveal/js/reveal.js"></script>
<script src="../reveal/js/jquery-3.5.0.min.js"></script>
<script>
Reveal.initialize({
width: "90%",
height: "90%",
slideNumber: 'c',
transition: 'none',
progress: true,
hash: true,
center: false,
dependencies: [
{ src: '../reveal/plugin/markdown/marked.js' },
{ src: '../reveal/plugin/markdown/markdown.js' },
{ src: '../reveal/plugin/highlight/highlight.js' },
{ src: '../reveal/plugin/notes/notes.js', async: true }
]
});
var footer = $('#footer-container').html();
$('div.reveal').append(footer);
</script>
</body>
</html>