-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathsignals.html
184 lines (158 loc) · 10.9 KB
/
signals.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
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"
"http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
<html xmlns="http://www.w3.org/1999/xhtml" xml:lang="en"
lang="en" dir="ltr">
<head>
<meta http-equiv="Content-Type" content="text/html; charset=utf-8" />
<title>signals</title>
<link rel="stylesheet" media="all" type="text/css" href="all.css" />
<link rel="stylesheet" media="screen" type="text/css" href="screen.css" />
<link rel="stylesheet" media="print" type="text/css" href="print.css" />
<link rel="stylesheet" media="all" type="text/css" href="export.css" />
</head>
<body>
<div class="dokuwiki export">
<!-- TOC START -->
<div id="dw__toc" class="dw__toc">
<h3 class="toggle">Table of Contents</h3>
<div>
<ul class="toc">
<li class="level1"><div class="li"><a href="#signals">Signals</a></div>
<ul class="toc">
<li class="level2"><div class="li"><a href="#signals1">Signals</a></div></li>
<li class="level2"><div class="li"><a href="#to_read">To Read</a></div></li>
<li class="level2"><div class="li"><a href="#books_i_like">Books I Like</a></div></li>
<li class="level2"><div class="li"><a href="#books_i_will_like">Books I Will Like</a></div></li>
</ul></li>
</ul>
</div>
</div>
<!-- TOC END -->
<h1 class="sectionedit1" id="signals">Signals</h1>
<div class="level1">
</div>
<!-- EDIT{"target":"section","name":"Signals","hid":"signals","codeblockOffset":0,"secid":1,"range":"1-23"} -->
<h2 class="sectionedit2" id="signals1">Signals</h2>
<div class="level2">
</div>
<h4 id="intro">Intro</h4>
<div class="level4">
<ul>
<li class="level1"><div class="li"> Awesome set of slides and homeworks for intro to signals class, very applicable: <a href="http://people.ee.duke.edu/maxim/teaching/fall08/" class="urlextern" title="http://people.ee.duke.edu/maxim/teaching/fall08/" rel="ugc nofollow">http://people.ee.duke.edu/maxim/teaching/fall08/</a></div>
</li>
</ul>
</div>
<h4 id="simple_but_elegant_filters">"Simple" but Elegant Filters</h4>
<div class="level4">
<ul>
<li class="level1"><div class="li"> Alex Holland: Morphology filter, <a href="0media/holland.09.pdf" class="media mediafile mf_pdf" title="holland.09.pdf">Recursive Fourier Transform for Non-Uniformly Sampled Signals</a>, <a href="https://www.google.com/search?tbm=pts&hl=en&q=An+improved+fault+tolerant+hard+disk+array+controller&gws_rd=ssl" class="urlextern" title="https://www.google.com/search?tbm=pts&hl=en&q=An+improved+fault+tolerant+hard+disk+array+controller&gws_rd=ssl" rel="ugc nofollow">Patent for scheduling on disk array systems</a> (particularly jump to his simplex optimization and cost function)</div>
</li>
<li class="level1"><div class="li"> <a href="0media/yelderman_cco.pdf" class="media mediafile mf_pdf" title="yelderman_cco.pdf">Continuous Cardiac Output algorithm</a></div>
</li>
</ul>
</div>
<h4 id="machine_learning_for_signal_processing">Machine Learning for Signal Processing</h4>
<div class="level4">
<ul>
<li class="level1 node"><div class="li"> <a href="http://courses.engr.illinois.edu/cs598ps/CS598PS/Main.html" class="urlextern" title="http://courses.engr.illinois.edu/cs598ps/CS598PS/Main.html" rel="ugc nofollow">Paris Smaragdis' Class at UIUC</a></div>
<ul>
<li class="level3"><div class="li"> Fun examples. Including figuring out nonlinear mappings of data (even into a jelly roll). Looks really difficult… </div>
</li>
<li class="level3"><div class="li"> His mapping of distances to US cities (<em>how did he do that??</em>) is really similar to Gerritsen's “apply weighting by charge repelling and spring attracting” in her <a href="http://youtu.be/s6p864XVxeU?t=14m25s" class="urlextern" title="http://youtu.be/s6p864XVxeU?t=14m25s" rel="ugc nofollow">video</a>.</div>
</li>
<li class="level3"><div class="li"> He really likes Non-Negative Matrix Factorization…and it seems for good reason! <a href="0media/smaragdis-waspaa2013keynote.pdf" class="media mediafile mf_pdf" title="smaragdis-waspaa2013keynote.pdf">smaragdis-waspaa2013keynote.pdf</a> (view in Adobe Reader to get flash player). Another paper that uses it: <a href="0media/smaragdis-spm2014.pdf" class="media mediafile mf_pdf" title="smaragdis-spm2014.pdf">smaragdis-spm2014.pdf</a></div>
</li>
</ul>
</li>
<li class="level1 node"><div class="li"> <a href="http://mlsp.cs.cmu.edu/courses/fall2012" class="urlextern" title="http://mlsp.cs.cmu.edu/courses/fall2012" rel="ugc nofollow">Bhiksha Raj's class at CMU</a></div>
<ul>
<li class="level2"><div class="li"> Excellent example of Expectation Maximization using looking at dice rolls. Makes much more intuitive sense than book explanations.</div>
</li>
</ul>
</li>
</ul>
</div>
<h4 id="pattern_recognition">Pattern Recognition</h4>
<div class="level4">
<ul>
<li class="level1 node"><div class="li"> Good slides on MACE filters: <a href="http://www.wvu.edu/~facerecognition/2D%20Matching%20Rep/VJ%20Kumar.pdf" class="urlextern" title="http://www.wvu.edu/~facerecognition/2D%20Matching%20Rep/VJ%20Kumar.pdf" rel="ugc nofollow">http://www.wvu.edu/~facerecognition/2D%20Matching%20Rep/VJ%20Kumar.pdf</a>. The original paper: <a href="http://www.cedar.buffalo.edu/~govind/CSE717/papers/Face%20Verification%20using%20Correlation%20Filters.pdf" class="urlextern" title="http://www.cedar.buffalo.edu/~govind/CSE717/papers/Face%20Verification%20using%20Correlation%20Filters.pdf" rel="ugc nofollow">http://www.cedar.buffalo.edu/~govind/CSE717/papers/Face%20Verification%20using%20Correlation%20Filters.pdf</a></div>
<ul>
<li class="level2"><div class="li"> They're really cool. Remember those bumps you see with correlation filters? It turns 'em into spikes! (the result is similar to doing a high pass filter, except a lot more official)</div>
</li>
</ul>
</li>
</ul>
</div>
<h4 id="time_series_analysis">Time Series Analysis</h4>
<div class="level4">
<ul>
<li class="level1"><div class="li"> Also known as real-time trading, identification, etc</div>
</li>
<li class="level1 node"><div class="li"> Great overview set of slides: <a href="http://alumni.cs.ucr.edu/~mvlachos/ICDM06/" class="urlextern" title="http://alumni.cs.ucr.edu/~mvlachos/ICDM06/" rel="ugc nofollow">http://alumni.cs.ucr.edu/~mvlachos/ICDM06/</a>. Talks about dynamic time warping, lcss algorithm, pca, all vs correlation and talks about benefits and disadvantages.</div>
<ul>
<li class="level2"><div class="li"> Anything you do other than correlation will be more expensive to compute, but could “see” your signal a lot better in the midst of noise / time dilation, etc.</div>
</li>
<li class="level2"><div class="li"> However, <code>“The more complicated the algorithm, the less it works”</code> –Bhiksha Raj <img src="/lib/images/smileys/icon_wink.gif" class="icon" alt=";-)" /></div>
</li>
</ul>
</li>
</ul>
</div>
<!-- EDIT{"target":"section","name":"Signals","hid":"signals1","codeblockOffset":0,"secid":2,"range":"24-2557"} -->
<h2 class="sectionedit3" id="to_read">To Read</h2>
<div class="level2">
<p>
<a href="0media/wt7.gif" class="media" title="wt7.gif"><img src="0media/wt7.gif" class="mediaright" align="right" alt="" width="100" /></a>
</p>
<ul>
<li class="level1"><div class="li"> <a href="http://users.rowan.edu/~polikar/WAVELETS/WTtutorial.html" class="urlextern" title="http://users.rowan.edu/~polikar/WAVELETS/WTtutorial.html" rel="ugc nofollow">Robi Polikar's Tutorial on Wavelet Analysis</a>. Still not quite sure what it can do, will need to ask Matt if it will be useful. Seems like it might be useful for data compression. Also has really good tutorial PDF's on Pattern Recognition (hidden markov models, nearest neighbor analysis) and Ensemble Decision Making.</div>
</li>
<li class="level1"><div class="li"> <a href="http://www.robots.ox.ac.uk/~parg/mlrg/papers/wavelets.pdf" class="urlextern" title="http://www.robots.ox.ac.uk/~parg/mlrg/papers/wavelets.pdf" rel="ugc nofollow">Wavelet Intro Paper</a></div>
</li>
<li class="level1"><div class="li"> <a href="0media/kavcic_correlsensitiveadapt.pdf" class="media mediafile mf_pdf" title="kavcic_correlsensitiveadapt.pdf">kavcic_correlsensitiveadapt.pdf</a> Correlation sensitive noise removal. </div>
</li>
</ul>
</div>
<!-- EDIT{"target":"section","name":"To Read","hid":"to_read","codeblockOffset":0,"secid":3,"range":"2558-3150"} -->
<h2 class="sectionedit4" id="books_i_like">Books I Like</h2>
<div class="level2">
</div>
<h4 id="understanding_dsp_by_lyons">Understanding DSP (by Lyons)</h4>
<div class="level4">
<p>
<a href="0media/home/9780137027415.jpg" class="media" title="home:9780137027415.jpg"><img src="0media/home/9780137027415.jpg" class="media" alt="" width="150" /></a>
</p>
</div>
<h5 id="chapters_1-3">Chapters 1-3</h5>
<div class="level5">
<ul>
<li class="level1"><div class="li"> Periodic sampling of 250 Hz allows a 10 Hz signal to look like a -10 Hz signal and a 240 and 260 Hz signal in the frequency domain! Doesn't really matter in Matlab experimenting, but can screw you up when you are demodulating a signal or getting high frequencies corrupting your signal (because they too can be interpreted as a slow-moving sine wave).</div>
</li>
<li class="level1"><div class="li"> Little difference between Hanning and Hamming windows, but you do want to use them! They reduce sidelobe leakage by a ton, but come at the cost of a 2X larger main lobe size. </div>
</li>
</ul>
</div>
<h5 id="fft_ch_4">FFT (Ch. 4)</h5>
<div class="level5">
<ul>
<li class="level1"><div class="li"> Skipping for now…maybe I'll get more interested eventually. Basically, there are redundancies in computing the DFT (can halve spectrum computation w/ real input signals), and some other things I don't understand yet.</div>
</li>
</ul>
</div>
<!-- EDIT{"target":"section","name":"Books I Like","hid":"books_i_like","codeblockOffset":0,"secid":4,"range":"3151-4045"} -->
<h2 class="sectionedit5" id="books_i_will_like">Books I Will Like</h2>
<div class="level2">
<ul>
<li class="level1 node"><div class="li"> <a href="http://www.dspguide.com/" class="urlextern" title="http://www.dspguide.com/" rel="ugc nofollow">Scientists and Engineers Guide to DSP</a> (and many other topics!)</div>
<ul>
<li class="level2"><div class="li"> Good overview on topics. Talks about compression, image / audio processing, and Benford's Law :)</div>
</li>
</ul>
</li>
</ul>
<p>
<a href="0media/bg-book.jpg" class="media" title="bg-book.jpg"><img src="0media/bg-book.jpg" class="media" alt="" width="200" /></a>
</p>
</div>
<!-- EDIT{"target":"section","name":"Books I Will Like","hid":"books_i_will_like","codeblockOffset":0,"secid":5,"range":"4046-"} --></div></body>
</html>