Skip to content

Commit d41d163

Browse files
pieroitpieroit
committed
add spark MLlib
1 parent 1ad2b01 commit d41d163

File tree

4 files changed

+249
-1
lines changed

4 files changed

+249
-1
lines changed

‎img/logo-spark.png

32 KB
Loading

‎img/spark_MLlib.png

21.4 KB
Loading

‎index.html

+2-1
Original file line numberDiff line numberDiff line change
@@ -48,7 +48,8 @@
4848
<h1>Machine Learning</h1>
4949
<h2><a href='introduction.html'>Introduction</a></h2>
5050
<h2><a href='applications.html'>Applications</a></h2>
51-
<h3><a href='javascript.html'>Javascript</a> &middot; <a href='wordpress.html'>WordPress</a></h3>
51+
<h3>
52+
<a href='javascript.html'>Javascript</a> &middot; <a href='wordpress.html'>WordPress</a> &middot; <a href='spark.html'>Spark</a></h3>
5253

5354
<p>
5455
<a rel="license" href="http://creativecommons.org/licenses/by-sa/4.0/">

‎spark.html

+247
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,247 @@
1+
<!doctype html>
2+
<html lang="en">
3+
4+
<head>
5+
<meta charset="utf-8">
6+
7+
<title>Machine Learning - Javascript</title>
8+
9+
<meta name="description" content="Machine learning in Javascript">
10+
<meta name="author" content="Piero Savastano">
11+
12+
<meta name="apple-mobile-web-app-capable" content="yes">
13+
<meta name="apple-mobile-web-app-status-bar-style" content="black-translucent">
14+
15+
<meta name="viewport" content="width=device-width, initial-scale=1.0, maximum-scale=1.0, user-scalable=no, minimal-ui">
16+
17+
<link rel="icon" type="image/x-icon" href="favicon.ico">
18+
19+
<link rel="stylesheet" href="css/reveal.css">
20+
<link rel="stylesheet" href="css/theme/moon.css" id="theme">
21+
<link rel="stylesheet" href="css/pieroit.css">
22+
23+
<!-- Code syntax highlighting -->
24+
<link rel="stylesheet" href="lib/css/zenburn.css">
25+
26+
<!-- Printing and PDF exports -->
27+
<script>
28+
var link = document.createElement( 'link' );
29+
link.rel = 'stylesheet';
30+
link.type = 'text/css';
31+
link.href = window.location.search.match( /print-pdf/gi ) ? 'css/print/pdf.css' : 'css/print/paper.css';
32+
document.getElementsByTagName( 'head' )[0].appendChild( link );
33+
</script>
34+
35+
<!--[if lt IE 9]>
36+
<script src="lib/js/html5shiv.js"></script>
37+
<![endif]-->
38+
</head>
39+
40+
<body>
41+
42+
<div class="reveal">
43+
44+
<!-- Any section element inside of this container is displayed as a slide -->
45+
<div class="slides">
46+
47+
<!-- Intro -->
48+
<section>
49+
<h1><a href="index.html">Machine Learning</a></h1>
50+
<div class="white">
51+
<img width="30%" src="img/logo-spark.png" />
52+
</div>
53+
<br/>
54+
<ul>
55+
<li>MLlib overview</li>
56+
<li>XXXXXXXXXX</li>
57+
</ul>
58+
</section>
59+
60+
<!-- Panorama -->
61+
<section>
62+
<a target="_blank" href="http://spark.apache.org/docs/latest/mllib-guide.html">
63+
<img width="30%" src="img/spark_MLlib.png" />
64+
</a>
65+
</section>
66+
67+
<!-- Demo -->
68+
<section>
69+
<section>
70+
<a href="https://github.com/NaturalNode/natural">
71+
<img width="100%" src="img/naturalJS.png" />
72+
</a>
73+
</section>
74+
<section>
75+
<pre class="stretch"><code class="hljs" data-trim>
76+
var natural = require('natural'),
77+
classifier = new natural.BayesClassifier();
78+
79+
80+
classifier.addDocument('my unit-tests failed.', 'software');
81+
classifier.addDocument('tried the program, but it was buggy.', 'software');
82+
classifier.addDocument('the drive has a 2TB capacity.', 'hardware');
83+
classifier.addDocument('i need a new power supply.', 'hardware');
84+
85+
classifier.train();
86+
87+
classifier.save('classifier.json', function(err, classifier) {
88+
// the classifier is saved to the classifier.json file!
89+
90+
natural.BayesClassifier.load('classifier.json', null, function(err, classifier) {
91+
console.log(classifier.classify('did you buy a new drive?'));
92+
});
93+
});
94+
</code></pre>
95+
</section>
96+
</section>
97+
98+
<section>
99+
<section>
100+
<a href="http://cs.stanford.edu/people/karpathy/convnetjs/">
101+
<img width="100%" src="img/convnetjs.png" />
102+
</a>
103+
</section>
104+
<section>
105+
<div>"The Deep Learning Conspiracy"</div>
106+
<a href="http://www.kdnuggets.com/wp-content/uploads/photo.jpg">
107+
<img width="100%" src="img/deep_learning_conspiracy.jpg" />
108+
</a>
109+
</section>
110+
</section>
111+
112+
<!-- Statistics libs -->
113+
<section>
114+
<a href="http://simplestatistics.org/">
115+
<img width="100%" src="img/simplestatisticsjs.png" />
116+
</a>
117+
</section>
118+
<section>
119+
<a href="http://jstat.github.io/">
120+
<img width="100%" src="img/jstat.png" />
121+
</a>
122+
</section>
123+
124+
<section>
125+
<h2>Takeaway</h2>
126+
<ul>
127+
<li>Algorithm vs. Model</li>
128+
<li>Javascript is good to run the Model</li>
129+
</ul>
130+
</section>
131+
132+
<!-- Example -->
133+
<section data-background="img/graph_viz.png" class="white">
134+
<h2>Let's write a model</h2>
135+
<ul>
136+
<li>Data gathering and cleaning</li>
137+
<li>Algorithm choice</li>
138+
<li>Model training</li>
139+
<li>Recommend</li>
140+
</ul>
141+
</section>
142+
143+
<section>
144+
<h2>Get data</h2>
145+
<div>Italian food</div>
146+
<pre><code class="hljs" data-trim>
147+
SELECT DISTINCT ?id ?label ?content WHERE {
148+
?id a dbo:Food .
149+
?id dbo:origin dbr:Italy .
150+
?id dbo:abstract ?content .
151+
?id rdfs:label ?label .
152+
153+
FILTER(LANG(?label) = "en")
154+
FILTER(LANG(?content) = "en")
155+
} LIMIT 10000
156+
</code></pre>
157+
<a href="http://dbpedia.org/sparql">
158+
Run on dbPedia
159+
</a>
160+
</section>
161+
162+
<!-- Measuring similarity between texts -->
163+
<section>
164+
<section>
165+
<h2>Algorithm</h2>
166+
<div>Jaccard/Tanimoto index</div>
167+
<img src="img/tanimoto.png" />
168+
</section>
169+
<section>
170+
<h2>Algorithm</h2>
171+
<div>Jaccard/Tanimoto index</div>
172+
<pre><code class="hljs" data-trim>
173+
function similarity( bagA, bagB ) {
174+
175+
var nIntersection = _.intersection( bagA, bagB).length;
176+
var nUnion = _.union( bagA, bagB ).length;
177+
178+
return nIntersection / nUnion;
179+
};
180+
</code></pre>
181+
</section>
182+
</section>
183+
184+
<!-- Build similarity graph -->
185+
<section>
186+
<h2>Similarity graph</h2>
187+
<span>
188+
<img height="200px" src="img/similarity_graph.jpg" />
189+
</span>
190+
<pre><code class="hljs" data-trim contenteditable>
191+
var similarityGraph = {
192+
"A":{
193+
"B": 0.1,
194+
"C": 0.8,
195+
},
196+
"B":{
197+
"A": 0.1,
198+
"C": 0.5,
199+
},
200+
"C":{
201+
"A": 0.8,
202+
"B": 0.5,
203+
}
204+
};
205+
</code></pre>
206+
</section>
207+
208+
<!-- Demo -->
209+
<section>
210+
<img width="800px" src="img/demo.jpg" />
211+
<div>
212+
<a href="https://github.com/pieroit/machine-learning-open-course/tree/gh-pages/nodejs-example">Show me the code!</a>
213+
</div>
214+
</section>
215+
216+
</div>
217+
218+
<script src="lib/js/head.min.js"></script>
219+
<script src="js/reveal.js"></script>
220+
221+
<script>
222+
223+
// Full list of configuration options available at:
224+
// https://github.com/hakimel/reveal.js#configuration
225+
Reveal.initialize({
226+
controls: true,
227+
progress: true,
228+
history: true,
229+
center: true,
230+
231+
transition: 'slide', // none/fade/slide/convex/concave/zoom
232+
233+
// Optional reveal.js plugins
234+
dependencies: [
235+
{ src: 'lib/js/classList.js', condition: function() { return !document.body.classList; } },
236+
{ src: 'plugin/markdown/marked.js', condition: function() { return !!document.querySelector( '[data-markdown]' ); } },
237+
{ src: 'plugin/markdown/markdown.js', condition: function() { return !!document.querySelector( '[data-markdown]' ); } },
238+
{ src: 'plugin/highlight/highlight.js', async: true, callback: function() { hljs.initHighlightingOnLoad(); } },
239+
{ src: 'plugin/zoom-js/zoom.js', async: true },
240+
{ src: 'plugin/notes/notes.js', async: true }
241+
]
242+
});
243+
244+
</script>
245+
246+
</body>
247+
</html>

0 commit comments

Comments
 (0)