forked from scikit-learn/scikit-learn.github.io
-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathscaling_strategies.html
410 lines (376 loc) · 23.2 KB
/
scaling_strategies.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
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
<!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">
<head>
<meta http-equiv="Content-Type" content="text/html; charset=utf-8" />
<title>6. Strategies to scale computationally: bigger data — scikit-learn 0.20.dev0 documentation</title>
<!-- htmltitle is before nature.css - we use this hack to load bootstrap first -->
<meta name="viewport" content="width=device-width, initial-scale=1.0" />
<link rel="stylesheet" href="../_static/css/bootstrap.min.css" media="screen" />
<link rel="stylesheet" href="../_static/css/bootstrap-responsive.css"/>
<link rel="stylesheet" href="../_static/nature.css" type="text/css" />
<link rel="stylesheet" href="../_static/pygments.css" type="text/css" />
<link rel="stylesheet" href="../_static/gallery.css" type="text/css" />
<script type="text/javascript">
var DOCUMENTATION_OPTIONS = {
URL_ROOT: '../',
VERSION: '0.20.dev0',
COLLAPSE_INDEX: false,
FILE_SUFFIX: '.html',
HAS_SOURCE: true,
SOURCELINK_SUFFIX: '.txt'
};
</script>
<script type="text/javascript" src="../_static/jquery.js"></script>
<script type="text/javascript" src="../_static/underscore.js"></script>
<script type="text/javascript" src="../_static/doctools.js"></script>
<script type="text/javascript" src="../_static/js/copybutton.js"></script>
<script type="text/javascript" src="../_static/js/extra.js"></script>
<script type="text/javascript" src="https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.0/MathJax.js?config=TeX-AMS_SVG"></script>
<link rel="shortcut icon" href="../_static/favicon.ico"/>
<link rel="author" title="About these documents" href="../about.html" />
<link rel="search" title="Search" href="../search.html" />
<link rel="next" title="7. Computational Performance" href="computational_performance.html" />
<link rel="prev" title="5.6.7. Kddcup 99 dataset" href="../datasets/kddcup99.html" />
<meta name="viewport" content="width=device-width, initial-scale=1.0" />
<script src="../_static/js/bootstrap.min.js" type="text/javascript"></script>
<script>
VERSION_SUBDIR = (function(groups) {
return groups ? groups[1] : null;
})(location.href.match(/^https?:\/\/scikit-learn.org\/([^\/]+)/));
</script>
<link rel="canonical" href="http://scikit-learn.org/stable/modules/scaling_strategies.html" />
<script type="text/javascript">
$("div.buttonNext, div.buttonPrevious").hover(
function () {
$(this).css('background-color', '#FF9C34');
},
function () {
$(this).css('background-color', '#A7D6E2');
}
);
function showMenu() {
var topNav = document.getElementById("scikit-navbar");
if (topNav.className === "navbar") {
topNav.className += " responsive";
} else {
topNav.className = "navbar";
}
};
</script>
</head>
<body>
<div class="header-wrapper">
<div class="header">
<p class="logo"><a href="../index.html">
<img src="../_static/scikit-learn-logo-small.png" alt="Logo"/>
</a>
</p><div class="navbar" id="scikit-navbar">
<ul>
<li><a href="../index.html">Home</a></li>
<li><a href="../install.html">Installation</a></li>
<li class="btn-li"><div class="btn-group">
<a href="../documentation.html">Documentation</a>
<a class="btn dropdown-toggle" data-toggle="dropdown">
<span class="caret"></span>
</a>
<ul class="dropdown-menu">
<li class="link-title">Scikit-learn <script>document.write(DOCUMENTATION_OPTIONS.VERSION + (VERSION_SUBDIR ? " (" + VERSION_SUBDIR + ")" : ""));</script></li>
<li><a href="../tutorial/index.html">Tutorials</a></li>
<li><a href="../user_guide.html">User guide</a></li>
<li><a href="classes.html">API</a></li>
<li><a href="../glossary.html">Glossary</a></li>
<li><a href="../faq.html">FAQ</a></li>
<li><a href="../developers/contributing.html">Contributing</a></li>
<li class="divider"></li>
<script>if (VERSION_SUBDIR != "stable") document.write('<li><a href="http://scikit-learn.org/stable/documentation.html">Stable version</a></li>')</script>
<script>if (VERSION_SUBDIR != "dev") document.write('<li><a href="http://scikit-learn.org/dev/documentation.html">Development version</a></li>')</script>
<li><a href="http://scikit-learn.org/dev/versions.html">All available versions</a></li>
<li><a href="../_downloads/scikit-learn-docs.pdf">PDF documentation</a></li>
</ul>
</div>
</li>
<li><a href="../auto_examples/index.html">Examples</a></li>
</ul>
<a href="javascript:void(0);" onclick="showMenu()">
<div class="nav-icon">
<div class="hamburger-line"></div>
<div class="hamburger-line"></div>
<div class="hamburger-line"></div>
</div>
</a>
<div class="search_form">
<div class="gcse-search" id="cse" style="width: 100%;"></div>
</div>
</div> <!-- end navbar --></div>
</div>
<!-- GitHub "fork me" ribbon -->
<a href="https://github.com/scikit-learn/scikit-learn">
<img class="fork-me"
style="position: absolute; top: 0; right: 0; border: 0;"
src="../_static/img/forkme.png"
alt="Fork me on GitHub" />
</a>
<div class="content-wrapper">
<div class="sphinxsidebar">
<div class="sphinxsidebarwrapper">
<div class="rel">
<div class="rellink">
<a href="../datasets/kddcup99.html"
accesskey="P">Previous
<br/>
<span class="smallrellink">
5.6.7. Kddcup...
</span>
<span class="hiddenrellink">
5.6.7. Kddcup 99 dataset
</span>
</a>
</div>
<div class="spacer">
</div>
<div class="rellink">
<a href="computational_performance.html"
accesskey="N">Next
<br/>
<span class="smallrellink">
7. Computatio...
</span>
<span class="hiddenrellink">
7. Computational Performance
</span>
</a>
</div>
<!-- Ad a link to the 'up' page -->
<div class="spacer">
</div>
<div class="rellink">
<a href="../user_guide.html">
Up
<br/>
<span class="smallrellink">
User Guide
</span>
<span class="hiddenrellink">
User Guide
</span>
</a>
</div>
</div>
<p class="doc-version"><b>scikit-learn v0.20.dev0</b><br/>
<a href="http://scikit-learn.org/stable/support.html#documentation-resources">Other versions</a></p>
<p class="citing">Please <b><a href="../about.html#citing-scikit-learn" style="font-size: 110%;">cite us </a></b>if you use the software.</p>
<ul>
<li><a class="reference internal" href="#">6. Strategies to scale computationally: bigger data</a><ul>
<li><a class="reference internal" href="#scaling-with-instances-using-out-of-core-learning">6.1. Scaling with instances using out-of-core learning</a><ul>
<li><a class="reference internal" href="#streaming-instances">6.1.1. Streaming instances</a></li>
<li><a class="reference internal" href="#extracting-features">6.1.2. Extracting features</a></li>
<li><a class="reference internal" href="#incremental-learning">6.1.3. Incremental learning</a></li>
<li><a class="reference internal" href="#examples">6.1.4. Examples</a></li>
<li><a class="reference internal" href="#notes">6.1.5. Notes</a></li>
</ul>
</li>
</ul>
</li>
</ul>
</div>
</div>
<input type="checkbox" id="nav-trigger" class="nav-trigger" checked />
<label for="nav-trigger"></label>
<div class="content">
<div class="documentwrapper">
<div class="bodywrapper">
<div class="body" role="main">
<div class="section" id="strategies-to-scale-computationally-bigger-data">
<span id="scaling-strategies"></span><h1>6. Strategies to scale computationally: bigger data<a class="headerlink" href="#strategies-to-scale-computationally-bigger-data" title="Permalink to this headline">¶</a></h1>
<p>For some applications the amount of examples, features (or both) and/or the
speed at which they need to be processed are challenging for traditional
approaches. In these cases scikit-learn has a number of options you can
consider to make your system scale.</p>
<div class="section" id="scaling-with-instances-using-out-of-core-learning">
<h2>6.1. Scaling with instances using out-of-core learning<a class="headerlink" href="#scaling-with-instances-using-out-of-core-learning" title="Permalink to this headline">¶</a></h2>
<p>Out-of-core (or “external memory”) learning is a technique used to learn from
data that cannot fit in a computer’s main memory (RAM).</p>
<p>Here is sketch of a system designed to achieve this goal:</p>
<blockquote>
<div><ol class="arabic simple">
<li>a way to stream instances</li>
<li>a way to extract features from instances</li>
<li>an incremental algorithm</li>
</ol>
</div></blockquote>
<div class="section" id="streaming-instances">
<h3>6.1.1. Streaming instances<a class="headerlink" href="#streaming-instances" title="Permalink to this headline">¶</a></h3>
<p>Basically, 1. may be a reader that yields instances from files on a
hard drive, a database, from a network stream etc. However,
details on how to achieve this are beyond the scope of this documentation.</p>
</div>
<div class="section" id="extracting-features">
<h3>6.1.2. Extracting features<a class="headerlink" href="#extracting-features" title="Permalink to this headline">¶</a></h3>
<p>2. could be any relevant way to extract features among the
different <a class="reference internal" href="feature_extraction.html#feature-extraction"><span class="std std-ref">feature extraction</span></a> methods supported by
scikit-learn. However, when working with data that needs vectorization and
where the set of features or values is not known in advance one should take
explicit care. A good example is text classification where unknown terms are
likely to be found during training. It is possible to use a stateful
vectorizer if making multiple passes over the data is reasonable from an
application point of view. Otherwise, one can turn up the difficulty by using
a stateless feature extractor. Currently the preferred way to do this is to
use the so-called <a class="reference internal" href="feature_extraction.html#feature-hashing"><span class="std std-ref">hashing trick</span></a> as implemented by
<a class="reference internal" href="generated/sklearn.feature_extraction.FeatureHasher.html#sklearn.feature_extraction.FeatureHasher" title="sklearn.feature_extraction.FeatureHasher"><code class="xref py py-class docutils literal"><span class="pre">sklearn.feature_extraction.FeatureHasher</span></code></a> for datasets with categorical
variables represented as list of Python dicts or
<a class="reference internal" href="generated/sklearn.feature_extraction.text.HashingVectorizer.html#sklearn.feature_extraction.text.HashingVectorizer" title="sklearn.feature_extraction.text.HashingVectorizer"><code class="xref py py-class docutils literal"><span class="pre">sklearn.feature_extraction.text.HashingVectorizer</span></code></a> for text documents.</p>
</div>
<div class="section" id="incremental-learning">
<h3>6.1.3. Incremental learning<a class="headerlink" href="#incremental-learning" title="Permalink to this headline">¶</a></h3>
<p>Finally, for 3. we have a number of options inside scikit-learn. Although not
all algorithms can learn incrementally (i.e. without seeing all the instances
at once), all estimators implementing the <code class="docutils literal"><span class="pre">partial_fit</span></code> API are candidates.
Actually, the ability to learn incrementally from a mini-batch of instances
(sometimes called “online learning”) is key to out-of-core learning as it
guarantees that at any given time there will be only a small amount of
instances in the main memory. Choosing a good size for the mini-batch that
balances relevancy and memory footprint could involve some tuning <a class="footnote-reference" href="#id2" id="id1">[1]</a>.</p>
<p>Here is a list of incremental estimators for different tasks:</p>
<blockquote>
<div><ul class="simple">
<li><dl class="first docutils">
<dt>Classification</dt>
<dd><ul class="first last">
<li><a class="reference internal" href="generated/sklearn.naive_bayes.MultinomialNB.html#sklearn.naive_bayes.MultinomialNB" title="sklearn.naive_bayes.MultinomialNB"><code class="xref py py-class docutils literal"><span class="pre">sklearn.naive_bayes.MultinomialNB</span></code></a></li>
<li><a class="reference internal" href="generated/sklearn.naive_bayes.BernoulliNB.html#sklearn.naive_bayes.BernoulliNB" title="sklearn.naive_bayes.BernoulliNB"><code class="xref py py-class docutils literal"><span class="pre">sklearn.naive_bayes.BernoulliNB</span></code></a></li>
<li><a class="reference internal" href="generated/sklearn.linear_model.Perceptron.html#sklearn.linear_model.Perceptron" title="sklearn.linear_model.Perceptron"><code class="xref py py-class docutils literal"><span class="pre">sklearn.linear_model.Perceptron</span></code></a></li>
<li><a class="reference internal" href="generated/sklearn.linear_model.SGDClassifier.html#sklearn.linear_model.SGDClassifier" title="sklearn.linear_model.SGDClassifier"><code class="xref py py-class docutils literal"><span class="pre">sklearn.linear_model.SGDClassifier</span></code></a></li>
<li><a class="reference internal" href="generated/sklearn.linear_model.PassiveAggressiveClassifier.html#sklearn.linear_model.PassiveAggressiveClassifier" title="sklearn.linear_model.PassiveAggressiveClassifier"><code class="xref py py-class docutils literal"><span class="pre">sklearn.linear_model.PassiveAggressiveClassifier</span></code></a></li>
<li><a class="reference internal" href="generated/sklearn.neural_network.MLPClassifier.html#sklearn.neural_network.MLPClassifier" title="sklearn.neural_network.MLPClassifier"><code class="xref py py-class docutils literal"><span class="pre">sklearn.neural_network.MLPClassifier</span></code></a></li>
</ul>
</dd>
</dl>
</li>
<li><dl class="first docutils">
<dt>Regression</dt>
<dd><ul class="first last">
<li><a class="reference internal" href="generated/sklearn.linear_model.SGDRegressor.html#sklearn.linear_model.SGDRegressor" title="sklearn.linear_model.SGDRegressor"><code class="xref py py-class docutils literal"><span class="pre">sklearn.linear_model.SGDRegressor</span></code></a></li>
<li><a class="reference internal" href="generated/sklearn.linear_model.PassiveAggressiveRegressor.html#sklearn.linear_model.PassiveAggressiveRegressor" title="sklearn.linear_model.PassiveAggressiveRegressor"><code class="xref py py-class docutils literal"><span class="pre">sklearn.linear_model.PassiveAggressiveRegressor</span></code></a></li>
<li><a class="reference internal" href="generated/sklearn.neural_network.MLPRegressor.html#sklearn.neural_network.MLPRegressor" title="sklearn.neural_network.MLPRegressor"><code class="xref py py-class docutils literal"><span class="pre">sklearn.neural_network.MLPRegressor</span></code></a></li>
</ul>
</dd>
</dl>
</li>
<li><dl class="first docutils">
<dt>Clustering</dt>
<dd><ul class="first last">
<li><a class="reference internal" href="generated/sklearn.cluster.MiniBatchKMeans.html#sklearn.cluster.MiniBatchKMeans" title="sklearn.cluster.MiniBatchKMeans"><code class="xref py py-class docutils literal"><span class="pre">sklearn.cluster.MiniBatchKMeans</span></code></a></li>
<li><a class="reference internal" href="generated/sklearn.cluster.Birch.html#sklearn.cluster.Birch" title="sklearn.cluster.Birch"><code class="xref py py-class docutils literal"><span class="pre">sklearn.cluster.Birch</span></code></a></li>
</ul>
</dd>
</dl>
</li>
<li><dl class="first docutils">
<dt>Decomposition / feature Extraction</dt>
<dd><ul class="first last">
<li><a class="reference internal" href="generated/sklearn.decomposition.MiniBatchDictionaryLearning.html#sklearn.decomposition.MiniBatchDictionaryLearning" title="sklearn.decomposition.MiniBatchDictionaryLearning"><code class="xref py py-class docutils literal"><span class="pre">sklearn.decomposition.MiniBatchDictionaryLearning</span></code></a></li>
<li><a class="reference internal" href="generated/sklearn.decomposition.IncrementalPCA.html#sklearn.decomposition.IncrementalPCA" title="sklearn.decomposition.IncrementalPCA"><code class="xref py py-class docutils literal"><span class="pre">sklearn.decomposition.IncrementalPCA</span></code></a></li>
<li><a class="reference internal" href="generated/sklearn.decomposition.LatentDirichletAllocation.html#sklearn.decomposition.LatentDirichletAllocation" title="sklearn.decomposition.LatentDirichletAllocation"><code class="xref py py-class docutils literal"><span class="pre">sklearn.decomposition.LatentDirichletAllocation</span></code></a></li>
</ul>
</dd>
</dl>
</li>
<li><dl class="first docutils">
<dt>Preprocessing</dt>
<dd><ul class="first last">
<li><a class="reference internal" href="generated/sklearn.preprocessing.StandardScaler.html#sklearn.preprocessing.StandardScaler" title="sklearn.preprocessing.StandardScaler"><code class="xref py py-class docutils literal"><span class="pre">sklearn.preprocessing.StandardScaler</span></code></a></li>
<li><a class="reference internal" href="generated/sklearn.preprocessing.MinMaxScaler.html#sklearn.preprocessing.MinMaxScaler" title="sklearn.preprocessing.MinMaxScaler"><code class="xref py py-class docutils literal"><span class="pre">sklearn.preprocessing.MinMaxScaler</span></code></a></li>
<li><a class="reference internal" href="generated/sklearn.preprocessing.MaxAbsScaler.html#sklearn.preprocessing.MaxAbsScaler" title="sklearn.preprocessing.MaxAbsScaler"><code class="xref py py-class docutils literal"><span class="pre">sklearn.preprocessing.MaxAbsScaler</span></code></a></li>
</ul>
</dd>
</dl>
</li>
</ul>
</div></blockquote>
<p>For classification, a somewhat important thing to note is that although a
stateless feature extraction routine may be able to cope with new/unseen
attributes, the incremental learner itself may be unable to cope with
new/unseen targets classes. In this case you have to pass all the possible
classes to the first <code class="docutils literal"><span class="pre">partial_fit</span></code> call using the <code class="docutils literal"><span class="pre">classes=</span></code> parameter.</p>
<p>Another aspect to consider when choosing a proper algorithm is that not all of
them put the same importance on each example over time. Namely, the
<code class="docutils literal"><span class="pre">Perceptron</span></code> is still sensitive to badly labeled examples even after many
examples whereas the <code class="docutils literal"><span class="pre">SGD*</span></code> and <code class="docutils literal"><span class="pre">PassiveAggressive*</span></code> families are more
robust to this kind of artifacts. Conversely, the latter also tend to give less
importance to remarkably different, yet properly labeled examples when they
come late in the stream as their learning rate decreases over time.</p>
</div>
<div class="section" id="examples">
<h3>6.1.4. Examples<a class="headerlink" href="#examples" title="Permalink to this headline">¶</a></h3>
<p>Finally, we have a full-fledged example of
<a class="reference internal" href="../auto_examples/applications/plot_out_of_core_classification.html#sphx-glr-auto-examples-applications-plot-out-of-core-classification-py"><span class="std std-ref">Out-of-core classification of text documents</span></a>. It is aimed at
providing a starting point for people wanting to build out-of-core learning
systems and demonstrates most of the notions discussed above.</p>
<p>Furthermore, it also shows the evolution of the performance of different
algorithms with the number of processed examples.</p>
<p class="centered">
<strong><a class="reference external" href="../auto_examples/applications/plot_out_of_core_classification.html"><img alt="accuracy_over_time" src="../_images/sphx_glr_plot_out_of_core_classification_0011.png" style="width: 512.0px; height: 384.0px;" /></a></strong></p><p>Now looking at the computation time of the different parts, we see that the
vectorization is much more expensive than learning itself. From the different
algorithms, <code class="docutils literal"><span class="pre">MultinomialNB</span></code> is the most expensive, but its overhead can be
mitigated by increasing the size of the mini-batches (exercise: change
<code class="docutils literal"><span class="pre">minibatch_size</span></code> to 100 and 10000 in the program and compare).</p>
<p class="centered">
<strong><a class="reference external" href="../auto_examples/applications/plot_out_of_core_classification.html"><img alt="computation_time" src="../_images/sphx_glr_plot_out_of_core_classification_0031.png" style="width: 512.0px; height: 384.0px;" /></a></strong></p></div>
<div class="section" id="notes">
<h3>6.1.5. Notes<a class="headerlink" href="#notes" title="Permalink to this headline">¶</a></h3>
<table class="docutils footnote" frame="void" id="id2" rules="none">
<colgroup><col class="label" /><col /></colgroup>
<tbody valign="top">
<tr><td class="label"><a class="fn-backref" href="#id1">[1]</a></td><td>Depending on the algorithm the mini-batch size can influence results or
not. SGD*, PassiveAggressive*, and discrete NaiveBayes are truly online
and are not affected by batch size. Conversely, MiniBatchKMeans
convergence rate is affected by the batch size. Also, its memory
footprint can vary dramatically with batch size.</td></tr>
</tbody>
</table>
</div>
</div>
</div>
</div>
</div>
</div>
<div class="clearer"></div>
</div>
</div>
<div class="footer">
© 2007 - 2018, scikit-learn developers (BSD License).
<a href="../_sources/modules/scaling_strategies.rst.txt" rel="nofollow">Show this page source</a>
</div>
<div class="rel">
<div class="buttonPrevious">
<a href="../datasets/kddcup99.html">Previous
</a>
</div>
<div class="buttonNext">
<a href="computational_performance.html">Next
</a>
</div>
</div>
<script type="text/javascript">
var _gaq = _gaq || [];
_gaq.push(['_setAccount', 'UA-22606712-2']);
_gaq.push(['_trackPageview']);
(function() {
var ga = document.createElement('script'); ga.type = 'text/javascript'; ga.async = true;
ga.src = ('https:' == document.location.protocol ? 'https://ssl' : 'http://www') + '.google-analytics.com/ga.js';
var s = document.getElementsByTagName('script')[0]; s.parentNode.insertBefore(ga, s);
})();
</script>
<script>
(function() {
var cx = '016639176250731907682:tjtqbvtvij0';
var gcse = document.createElement('script'); gcse.type = 'text/javascript'; gcse.async = true;
gcse.src = 'https://cse.google.com/cse.js?cx=' + cx;
var s = document.getElementsByTagName('script')[0]; s.parentNode.insertBefore(gcse, s);
})();
</script>
</body>
</html>