-
Notifications
You must be signed in to change notification settings - Fork 146
/
spark-mllib-ml-pipelines.html
279 lines (246 loc) · 11.9 KB
/
spark-mllib-ml-pipelines.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
<!doctype html>
<html lang="en">
<head>
<meta charset="utf-8">
<meta name="viewport" content="width=device-width, initial-scale=1.0, maximum-scale=1.0, user-scalable=no">
<title>Apache Spark™ Workshop | Spark MLlib | ML Pipelines</title>
<meta name="description" content="Apache Spark™ Workshop | Spark MLlib | ML Pipelines">
<meta name="author" content="Jacek Laskowski">
<link rel="stylesheet" href="reveal.js/css/reveal.css">
<link rel="stylesheet" href="reveal.js/css/theme/beige.css">
<!-- Theme used for syntax highlighting of code -->
<link rel="stylesheet" href="reveal.js/lib/css/zenburn.css">
<!-- Jacek: custom formatting -->
<link rel="stylesheet" href="revealjs-css/jacek.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.js/css/print/pdf.css' : 'reveal.js/css/print/paper.css';
document.getElementsByTagName('head')[0].appendChild(link);
</script>
</head>
<body>
<div class="reveal">
<div class="footer">
<footer style="font-size: small;">
© <a href="https://medium.com/@jaceklaskowski">Jacek Laskowski</a> 2020 / <a href="https://twitter.com/jaceklaskowski">@JacekLaskowski</a>
/ jacek@japila.pl
</footer>
</div>
<div class="slides">
<section class="intro" data-transition="zoom" id="home">
<p>
<img width="5%" style="background:none; border:none; box-shadow:none;" data-src="images/scala-logo.png">
<img width="17%" style="background:none; border:none; box-shadow:none;" data-src="images/spark-logo.png">
<img width="8%" src="images/jacek_laskowski_20141201_512px.png" style="border: 0">
</p>
<h1>ML Pipelines</h1>
<h3>Apache Spark 2.4 / Spark MLlib</h3>
<br>
<hr>
<h4 style="font-size: smaller;">
<a href="https://twitter.com/jaceklaskowski">@jaceklaskowski</a> / <a href="https://stackoverflow.com/users/1305344/jacek-laskowski">StackOverflow</a> / <a href="https://github.com/jaceklaskowski">GitHub</a> / <a href="https://www.linkedin.com/in/jaceklaskowski/">LinkedIn</a>
<br>
The "Internals" Books: <a href="https://books.japila.pl/apache-spark-internals">Apache Spark</a> / <a href="https://bit.ly/spark-sql-internals">Spark SQL</a> / <a href="https://bit.ly/spark-structured-streaming">Spark Structured Streaming</a> / <a href="https://books.japila.pl/delta-lake-internals">Delta Lake</a>
</h4>
</section>
<section id="agenda" data-markdown>
<textarea data-template>
## Agenda
1. [ML Pipelines (spark.ml)](#/pipelines)
1. [Transformers](#/transformers)
1. [Estimators](#/estimators)
1. [Models](#/models)
1. [Evaluators](#/evaluators)
1. [CrossValidator](#/crossvalidator)
1. [Persistence — MLWriter and MLReader](#/persistence)
1. [Example](#/example)
</textarea>
</section>
<section>
<section id="pipelines">
<h2>ML Pipelines (spark.ml)</h2>
<ol>
<li>DataFrame-based API under <b>spark.ml</b> package.
<pre><code>import org.apache.spark.ml._</code></pre>
<ul>
<li><b>spark.mllib</b> package obsolete (as of Spark 2.0)</li>
</ul>
</li>
<li>Switch to <a href="https://books.japila.pl/apache-spark-internals">The Internals Of Apache Spark</a>
<ul>
<li><a href="https://books.japila.pl/apache-spark-internals/apache-spark-internals/2.4.4/spark-mllib/spark-mllib-pipelines.html">ML Pipelines (spark.ml)</a></li>
</ul>
</li>
</ol>
</section>
<section>
<h2>text document pipeline</h2>
<img src="images/ml-Pipeline.png" alt="ML Pipeline" style="background:none; border:none; box-shadow:none;" />
<small>From <a href="http://spark.apache.org/docs/latest/ml-pipeline.html#how-it-works">the official documentation of Apache Spark</a></small>
</section>
</section>
<section id="transformers">
<h2>Transformers</h2>
<ol>
<li><b>Transformer</b> transforms DataFrame into "enhanced" DataFrame.
<pre><code>transformer: DataFrame =[transform]=> DataFrame</code></pre>
</li>
<li>Switch to <a href="https://books.japila.pl/apache-spark-internals">The Internals Of Apache Spark</a>
<ul>
<li><a href="https://books.japila.pl/apache-spark-internals/apache-spark-internals/2.4.4/spark-mllib/spark-mllib-transformers.html">Transformers</a></li>
</ul>
</li>
</ol>
</section>
<section id="estimators">
<h2>Estimators</h2>
<ol>
<li><b>Estimator</b> produces <b>Model</b> (Transformer) for training DataFrame
<pre><code>estimator: DataFrame =[fit]=> Model</code></pre>
</li>
<li>Switch to <a href="https://books.japila.pl/apache-spark-internals">The Internals Of Apache Spark</a>
<ul>
<li><a href="https://books.japila.pl/apache-spark-internals/apache-spark-internals/2.4.4/spark-mllib/spark-mllib-estimators.html">Estimators</a></li>
</ul>
</li>
</ol>
</section>
<section>
<section id="models">
<h2>Models</h2>
<ol>
<li><b>Model</b> - transformer that generates <b>predictions</b> for DataFrame
<pre><code>model: DataFrame =[predict]=> DataFrame (with predictions)</code></pre>
</li>
<li>Switch to <a href="https://books.japila.pl/apache-spark-internals">The Internals Of Apache Spark</a>
<ul>
<li><a href="https://books.japila.pl/apache-spark-internals/apache-spark-internals/2.4.4/spark-mllib/spark-mllib-models.html">ML Pipeline Models</a></li>
</ul>
</li>
</ol>
</section>
<section>
<h2>text document pipeline model</h2>
<img src="images/ml-PipelineModel.png" alt="ML Pipeline Model" style="background:none; border:none; box-shadow:none;" />
<small>From <a href="http://spark.apache.org/docs/latest/ml-pipeline.html#how-it-works">the official documentation of Apache Spark</a></small>
</section>
</section>
<section id="evaluators">
<h2>Evaluators</h2>
<ol>
<li><b>Evaluator</b> - transformation that <b>measures</b> effectiveness of Model, i.e. how good a model is.
<pre><code>evaluator: DataFrame =[evaluate]=> Double</code></pre>
</li>
<li>Switch to <a href="https://books.japila.pl/apache-spark-internals">The Internals Of Apache Spark</a>
<ul>
<li><a href="https://books.japila.pl/apache-spark-internals/apache-spark-internals/2.4.4/spark-mllib/spark-mllib-Evaluator.html">Evaluator</a></li>
</ul>
</li>
</ol>
</section>
<section id="crossvalidator">
<h2>CrossValidator</h2>
<ol>
<li><b>CrossValidator</b> - estimator that gives the best Model for parameters
<pre><code>import org.apache.spark.ml.tuning.CrossValidator</code></pre>
</li>
<li>Switch to <a href="https://books.japila.pl/apache-spark-internals">The Internals Of Apache Spark</a>
<ul>
<li><a href="https://books.japila.pl/apache-spark-internals/apache-spark-internals/2.4.4/spark-mllib/spark-mllib-CrossValidator.html">CrossValidator</a></li>
</ul>
</li>
</ol>
</section>
<section id="persistence">
<h2>Persistence — MLWriter and MLReader</h2>
<ol>
<li>Allows saving and loading models
<pre><code>model.write
.overwrite()
.save("/path/where/to/save/model")</code></pre>
<pre><code>val model =
PipelineModel.load("/path/with/model")</code></pre>
</li>
<li>Switch to <a href="https://books.japila.pl/apache-spark-internals">The Internals Of Apache Spark</a>
<ul>
<li><a href="https://books.japila.pl/apache-spark-internals/apache-spark-internals/2.4.4/spark-mllib/spark-mllib-pipelines-persistence.html">ML Persistence — Saving and Loading Models and Pipelines</a></li>
</ul>
</li>
</ol>
</section>
<section id="example">
<h2>Example</h2>
<ol>
<li>Switch to <a href="https://books.japila.pl/apache-spark-internals">The Internals Of Apache Spark</a>
<ul>
<li><a href="https://books.japila.pl/apache-spark-internals/apache-spark-internals/2.4.4/spark-mllib/spark-mllib-pipelines-example-classification.html">Example — Text Classification</a></li>
</ul>
</li>
</ol>
</section>
<section id="recap" data-markdown>
<textarea data-template>
## Recap
1. [ML Pipelines (spark.ml)](#/pipelines)
1. [Transformers](#/transformers)
1. [Estimators](#/estimators)
1. [Models](#/models)
1. [Evaluators](#/evaluators)
1. [CrossValidator](#/crossvalidator)
1. [Persistence — MLWriter and MLReader](#/persistence)
1. [Example](#/example)
</textarea>
</section>
<section style="text-align: left" data-markdown id="questions">
<textarea data-template>
# Questions?
* Follow [@jaceklaskowski](https://twitter.com/jaceklaskowski) on twitter (DMs open)
* Upvote my questions and answers on [StackOverflow](http://stackoverflow.com/users/1305344/jacek-laskowski)
* Contact me at **jacek@japila.pl**
* Connect with me at [LinkedIn](https://www.linkedin.com/in/jaceklaskowski/)
</textarea>
</section>
</div>
</div>
<script src="reveal.js/lib/js/head.min.js"></script>
<script src="reveal.js/js/reveal.js"></script>
<script>
// More info about config & dependencies:
// - https://github.com/hakimel/reveal.js#configuration
// - https://github.com/hakimel/reveal.js#dependencies
Reveal.initialize({
controls: true,
progress: true,
history: true,
center: true,
slideNumber: true,
transition: 'slide', // none/fade/slide/convex/concave/zoom
menu: {
markers: true,
openSlideNumber: true
},
dependencies: [
{ src: 'reveal.js/lib/js/classList.js', condition: function () { return !document.body.classList; } },
{ src: 'reveal.js/plugin/markdown/marked.js' },
{ src: 'reveal.js/plugin/markdown/markdown.js' },
{ src: 'reveal.js/plugin/zoom-js/zoom.js', async: true },
{ src: 'reveal.js/plugin/notes/notes.js', async: true },
{ src: 'reveal.js/plugin/highlight/highlight.js', async: true, callback: function () { hljs.initHighlightingOnLoad(); } }
]
});
</script>
<script>
(function (i, s, o, g, r, a, m) {
i['GoogleAnalyticsObject'] = r; i[r] = i[r] || function () {
(i[r].q = i[r].q || []).push(arguments)
}, i[r].l = 1 * new Date(); a = s.createElement(o),
m = s.getElementsByTagName(o)[0]; a.async = 1; a.src = g; m.parentNode.insertBefore(a, m)
})(window, document, 'script', '//www.google-analytics.com/analytics.js', 'ga');
ga('create', 'UA-45999426-3', 'auto');
ga('send', 'pageview');
</script>
</body>
</html>