-
Notifications
You must be signed in to change notification settings - Fork 0
Expand file tree
/
Copy pathindex.html
More file actions
660 lines (607 loc) · 28.5 KB
/
index.html
File metadata and controls
660 lines (607 loc) · 28.5 KB
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
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
<!DOCTYPE html>
<html lang="en">
<head>
<title>CS839 - Data Management for Machine Learning Applications (Spring 2018)</title>
<!-- HTML5 Shim and Respond.js IE8 support of HTML5 elements and media queries -->
<!-- WARNING: Respond.js doesn't work if you view the page via file:// -->
<!--[if lt IE 9]>
<script src="https://oss.maxcdn.com/html5shiv/3.7.2/html5shiv.min.js"></script>
<script src="https://oss.maxcdn.com/respond/1.4.2/respond.min.js"></script>
<![endif]-->
<link rel="stylesheet" href="https://maxcdn.bootstrapcdn.com/bootstrap/3.2.0/css/bootstrap.min.css">
<script src="https://ajax.googleapis.com/ajax/libs/jquery/2.1.1/jquery.min.js"></script>
<script src="https://maxcdn.bootstrapcdn.com/bootstrap/3.2.0/js/bootstrap.min.js"></script>
<link rel="stylesheet" href="./style.css">
</head>
<div class="container">
<div class="navbar navbar-default" role="navigation">
<div class="container-fluid">
<div class="navbar-header">
<button type="button" class="navbar-toggle collapsed" data-toggle="collapse" data-target=".navbar-collapse">
<span class="sr-only">Toggle navigation</span>
<span class="icon-bar"></span>
<span class="icon-bar"></span>
<span class="icon-bar"></span>
</button>
<a class="navbar-brand" href="#">CS 839</a>
</div>
<div class="collapse navbar-collapse" id="bs-example-navbar-collapse-1">
<ul class="nav navbar-nav">
<li class="dropdown">
<a href="#" class="dropdown-toggle" data-toggle="dropdown">About <span class="caret"></span></a>
<ul class="dropdown-menu" role="menu">
<li><a href="#description">Description</a></li>
<li><a href="#logistics">Class Logistics</a></li>
<li><a href="#lecture">Lecture Plan</a></li>
<li><a href="#grading">Grading</a></li>
<li class="divider"></li>
<li><a href="#late-policy">Late Policy</a></li>
</ul>
</li>
<li><a href="#office-hours">Office Hours</a></li>
<li><a href="https://piazza.com/wisc/spring2018/cs839">Piazza</a></li>
</ul>
</div>
</div>
</div>
<div class="jumbotron">
<h1>CS839 Data Management for Machine Learning Applications</h1>
<h3>COMP SCI 1257 on TuTh 2:30-3:45pm</h3>
<p class="lead">Chat with us on <a href="https://piazza.com/wisc/spring2018/cs839">the course Piazza site</a> if you have any questions!</p>
<p class="lead">Class repository: <a href="https://drive.google.com/drive/folders/1soJMN9FuuLOcC9diYtdSvfZZ9aZA4qBz?usp=sharing">CS839 Drive</a></p>
</div>
<div class="panel panel-default">
<div class="panel-heading">Description</div>
<div class="panel-body">
<a name="description"></a>
<p>We are seeing widespread investments in machine learning (ML) that enable computers to interpret what they see, communicate in natural language, answer complex questions, and interact with their environment. There is a hidden catch, however: all state-of-the-art ML systems rely on high-effort data management tasks like data exploration, data preparation and data cleaning. The goal of this seminar course is to study data management challenges that arise in the context of machine learning pipelines. The focus will be on cutting-edge problems in the context of ML pipelines, related to (1) data exploration and understanding, (2) data integration, cleaning, and validation, and (3) data preparation for ML models and serving of production ML applications.
The seminar will be very interactive and collaborative. The topics covered and the depth of coverage will depend on the participants' input and interests.
The goal of the course is to give you an indepth look at an important, emerging topic in data management research. This course will provide research opportunities in the areas of data management, human-computer interaction, and machine learning. Along the way, you will also pick up some practical experience in reading and presenting research papers, synthesizing research across desperate areas, using existing tools, and doing a course project that ideally will lead to a publishable paper.
</p>
</div>
</div>
<div class="panel panel-default">
<div class="panel-heading">Class Logistics</div>
<div class="panel-body">
<a name="logistics"></a>
<p> <b>Class Format</b>
<ul>
<li><p>This is a seminar course. Each class will consist of presentations and discussion. Students will be required to do a class project for the course (60%). A significant portion of the grade will be based on class participation, which includes paper presentations, contributions to paper reviews, and paper discussions (40%). Because of the interactive nature of the course, and space limitations, auditing is discouraged.
</p></li>
</ul>
</p>
<p> <b>Prerequisites</b>
<ul>
<li><p>Mathematical maturity and a basic course in probability required. Background in algorithms, databases, machine learning, and graphical models suggested.</p></li>
</ul>
</p>
<p><b>Assignments</b>
<ul>
<li>You will need to form <b> groups of up to three</b> people and work on certain assignments as described below.
<li>A research class project: Each group will work on a research project and file a single submission. The project will be broken down to five assignments: (1) initial research proposal, (2) Intermediate report, (3) final report, (4) final presentation (in class), and (5) poster presentation (people from the entire department will be invited).</li>
<li>Questions, Comments, and Responses (QCRs): During each class you will need to provide 3 questions and 3 comments for each (mandatory) paper. QCs will be <b>individual assignments</b>. For each class one of the groups (see above) will be responsible to lead a discussion on the papers and provide answers to all posted questions during class. Comments and answers will then be summarized in a written report which will be submitted and shared with everyone in the class. </li>
<li>There will be <b> no midterm or final exams</b>.
</ul>
</p>
<p><b>Misc</b>
<ul>
<li>Class time may be adjusted to accomodate external talks releated to the class.</li>
<li>Google drive for deliverables: <a href="https://drive.google.com/drive/folders/1soJMN9FuuLOcC9diYtdSvfZZ9aZA4qBz?usp=sharing">CS839 Drive</a></li>
</ul>
</p>
</div>
</div>
<div class="panel panel-default">
<div class="panel-heading">
<h2 class="panel-title">Tentative Lecture Plan (Subject to Change)</h2>
</div>
<div class="panel-body">
<a name="lecture"></a>
<p>
<hr />
<table class="plan">
<tbody>
<tr>
<th>#</th>
<th>Date</th>
<th>Topic</th>
<th>Lecture Materials</th>
<th>Reading Material</th>
<th>Assignments</th>
</tr>
<tr>
<td align="center" colspan="6" class="small-section-header">Introduction and Class Overview</td>
</tr>
<tr>
<td>1</td>
<td>1/23</td>
<td>Logistics and Data Management for Production ML</td>
<td class="side-borders">
<!-- Lecture -->
<a href=""><a href="lectures/Lecture_1_Intro.pptx">Lecture 1</a>
</td>
<td class="side-borders">
<ul>
<li>Kaggle: <a href="https://www.kaggle.com/surveys/2017">The state of data science</a></li>
<li>Polyzotis et al. <a href="papers/p1723-polyzotis.pdf">Data Management Challenges in Production Machine Learning</a>, SIGMOD 2017</li>
</ul>
</td>
<td></td>
</tr>
<tr>
<td>2</td>
<td>1/25</td>
<td>DB and ML integration: A systems percpective</td>
<td class="side-borders">
<!-- Lecture -->
<a href=""><a href="lectures/Lecture_2_DML.pptx">Lecture 2</a>
</td>
<td class="side-borders">
<ul>
<li><a href="https://databricks.com/session/inside-apache-systemml">Inside SystemML</a></li>
<li>Kumar et al. <a href="papers/p1717-kumar.pdf">Data Management in Machine Learning</a>, SIGMOD 2017</li>
<li><a href="papers/EECS-2017-159.pdf">System Challenges for AI</a></li>
</ul>
</td>
<td></td>
</tr>
<tr>
<td align="center" colspan="6" class="small-section-header">Data Exploration and Understanding</td>
</tr>
<tr>
<td>3</td>
<td>1/30</td>
<td>What to expect from data: Data driven visualizations</td>
<td class="side-borders">
<!-- Lecture -->
<a href=""><a href="lectures/Lecture_3_Vis.pptx">Lecture 3</a>
</td>
<td class="side-borders">
<ul>
<li>Vartak et al. <a href="http://www.vldb.org/pvldb/vol8/p2182-vartak.pdf">SeeDB</a>, VLDB 2015</li>
<li><a href="https://medium.com/@adityagp/enabling-data-science-for-the-majority-6255fabe230a">Enabling Data Science for the Majority</a> (Read but <b>no QCRs</b>)</li>
<li>Siddiqui et al. <a href ="http://www.vldb.org/pvldb/vol10/p457-siddiqui.pdf">ZenVisage</a>, VLDB 2017 <b>(Optional)</b></li>
<li>Gonzalez et al. <a href ="papers/p1061-gonzalez.pdf">Google Fusion Tables</a>, SIGMOD 2010 <b>(Optional)</b></li>
</ul>
</td>
<td>
</td>
</tr>
<tr>
<td>4</td>
<td>2/1</td>
<td>From data cubes to feature-based analysis</td>
<td class="side-borders">
<!-- Lecture -->
<a href=""><a href="lectures/Lecture_4_DD.pptx">Lecture 4</a>
</td>
<td class="side-borders">
<ul>
<li>Jogklekar et al. <a href="https://arxiv.org/pdf/1412.0364.pdf">Interactive Data Exploration with Smart Drill-Down</a>, ICDE 2016</li>
<li>Sathe and Sarawagi <a href="https://www.cse.iitb.ac.in/~sunita/papers/vldb01.pdf">Intelligent Rollups in Multidimensional OLAP Data</a>, VLDB 2001 <b>(Optional)</b></li>
<li>Harinarayan et al. <a href ="papers/p205-harinarayan.pdf">Implementing Data Cubes Efficiently</a>, SIGMOD 1996 <b>(Optional)</b></li>
<li>Kahng et al. <a href ="https://www.cc.gatech.edu/~dchau/papers/16-sigmod-mlcube.pdf">Visual Exploration of Machine Learning Results using Data Cube Analysis</a>, HILDA Workshop @ SIGMOD 2016 <b>(Optional)</b></li>
</ul>
</td>
<td></td>
</tr>
<tr>
<td>5</td>
<td>2/6</td>
<td>Leave no relevant data behind: Data search</td>
<td class="side-borders">
<!-- Lecture -->
<a href=""><a href="lectures/Lecture_5_DS.pptx">Lecture 5</a>
</td>
<td class="side-borders">
<ul>
<li>Halevy et al. <a href="https://static.googleusercontent.com/media/research.google.com/en//pubs/archive/45390.pdf">Goods: Organizing Google's Datasets</a>, SIGMOD 2016</li>
<li>Hellerstein et al. <a href="http://cidrdb.org/cidr2017/papers/p111-hellerstein-cidr17.pdf">Ground: A Data Context Service</a>, CIDR 2017 <b>(Optional)</b></li>
<li>Rekatsinas et al. <a href="http://web2.research.att.com/export/sites/att_labs/techdocs/TD_101504.pdf">Finding Quality in Quantity</a>, CIDR 2015 <b>(Optional)</b></li>
</ul>
</td>
<td></td>
</tr>
<tr>
<td>6</td>
<td>2/8</td>
<td>Data hubs: Version control for datasets</td>
<td class="side-borders">
<!-- Lecture -->
<a href=""><a href="lectures/Lecture_6_DV.pptx">Lecture 6</a>
</td>
<td class="side-borders">
<ul>
<li>Bhattacherjee et al. <a href="http://www.vldb.org/pvldb/vol8/p1346-bhattacherjee.pdf">Principles of Dataset Versioning</a>, VLDB 2015</li>
<li>Miao et al. <a href="https://arxiv.org/pdf/1610.04963.pdf">ProvDB: A System for Lifecycle Management of Collaborative Analysis Workflows</a>, HILDA Workshop@SIGMOD 2017 <b>(Optional)</b></li>
</ul>
</td>
<td></td>
</tr>
<tr>
<td align="center" colspan="6" class="small-section-header">Data Preparation: Extraction, Integration, Cleaning</td>
</tr>
<tr>
<td>7</td>
<td>2/13</td>
<td>Knowledge Base Construction: From dark data to insights</td>
<td class="side-borders">
<!-- Lecture -->
<a href=""><a href="lectures/Lecture_7_KBC.pptx">Lecture 7</a>
</td>
<td class="side-borders">
<ul>
<li>Shin et al. <a href="http://www.vldb.org/pvldb/vol8/p1310-shin.pdf">Incremental Knowledge Base Construction Using DeepDive</a>, VLDB 2015</li>
<li>Dong et al. <a href="https://static.googleusercontent.com/media/research.google.com/en//pubs/archive/45634.pdf">Knowledge Vault: A Web-Scale Approach to Probabilistic Knowledge Fusion</a>, KDD 2014 <b>(Optional)</b></li>
<li>Carlson et al. <a href="http://www.cs.cmu.edu/~acarlson/papers/carlson-aaai10.pdf">Toward an Architecture for Never-Ending Language Learning
</a>, AAAI 2010 <b>(Optional)</b></li>
</ul>
</td>
<td></td>
</tr>
<tr>
<td>8</td>
<td>2/15</td>
<td>No Class (Theo @ SysML)</td>
<td class="side-borders">
<!-- Lecture -->
</td>
<td class="side-borders"></td>
<td></td>
</tr>
<tr>
<td>9</td>
<td>2/20</td>
<td>Creating Training Data</td>
<td class="side-borders">
<!-- Lecture -->
<a href=""><a href="lectures/Lecture_8_WS.pptx">Lecture 8</a>
</td>
<td class="side-borders">
<ul>
<li>Ratner et al. <a href="http://www.vldb.org/pvldb/vol11/p269-ratner.pdf">Snorkel: Rapid Training Data Creation with Weak Supervision</a>, VLDB 2018</li>
<li><a href="https://dawn.cs.stanford.edu/2017/07/16/weak-supervision/">Weak Superivsion blog</a> (Read but <b>no QCRs</b>)</li>
<li>Ratner et al. <a href="https://dawn.cs.stanford.edu/pubs/snorkel-nips2016.pdf">Data Programming: Creating Large Training Sets, Quickly</a>, NIPS 2016 <b>(Optional)</b></li>
<li>Carlson et al. <a href="http://www.cs.cmu.edu/~acarlson/papers/carlson-aaai10.pdf">Toward an Architecture for Never-Ending Language Learning
</a>, AAAI 2010 <b>(Optional)</b></li>
</ul>
</td>
<td>Example class projects posted <a href="https://docs.google.com/document/d/1ORQnnjpMzjxBkJblsNoPXZM9xxNF6o4ZSIz5c-v6WsA/edit">here</a>!</td>
</tr>
<tr>
<td>10</td>
<td>2/22</td>
<td>Data Integration: Entity Resolution 1</td>
<td class="side-borders">
<!-- Lecture -->
<a href=""><a href="lectures/Lecture_9_ER1.pptx">Lecture 9</a>
</td>
<td class="side-borders">
<ul>
<li> Fellegi and Sunter <a href="papers/Fellegi69.pdf">A Theory for Record Linkage</a>, Journal of the American Statistical Association 1969</li>
<li> Getoor and Machanavajjhala <a href="https://www.umiacs.umd.edu/~getoor/Tutorials/ER_VLDB2012.pdf">Entity Resolution: Tutorial</a>, VLDB 2012 <b>no QCRs</b></li>
<li> Cohen et al. <a href="http://www.cs.cmu.edu/~wcohen/postscript/kdd-2003-match-ws.pdf">A Comparison of String Metrics for Matching Names and Records</a>, IJCAI 2003 <b>(Optional)</b></li>
<li> Singla and Domingos <a href="https://homes.cs.washington.edu/~pedrod/papers/icdm06.pdf">Entity Resolution with Markov Logic</a>, ICDM 2006 <b>(Optional)</b></li>
</ul>
</td>
<td></td>
</tr>
<tr>
<td>11</td>
<td>2/27</td>
<td>Data Integration: Entity Resolution 2</td>
<td class="side-borders">
<!-- Lecture -->
<a href=""><a href="lectures/magellan_talk_theo_class.pdf">Lecture 10</a>
</td>
<td class="side-borders">
<ul>
<li> Wang et al. <a href="http://www.vldb.org/pvldb/vol4/p622-wang.pdf">Entity Matching: How Similar is Similar</a>, VLDB 2011</li>
<li> Konda et al. <a href="http://www.vldb.org/pvldb/vol9/p1197-pkonda.pdf">Magellan: Toward Building Entity Matching Management Systems</a>, VLDB 2016 <b>(Optional)</b></li>
<li> Das et al. <a href="http://pages.cs.wisc.edu/~anhai/papers/falcon-tr.pdf">Falcon: Scaling Up Hands-Off Crowdsourced Entity Mathing to Build Cloud Services</a>, SIGMOD 2017 <b>(Optional)</b></li>
</ul>
</td>
<td><b>Project proposal due!</b></td>
</tr>
<tr>
<td>12</td>
<td>3/1</td>
<td>Data Integration: Data Fusion</td>
<td class="side-borders">
<!-- Lecture -->
<a href=""><a href="lectures/Lecture_11_DF.pdf">Lecture 11</a>
</td>
<td class="side-borders">
<ul>
<li> Rekatsinas et al. <a href="papers/p1399-rekatsinas.pdf">SLiMFast: Guaranteed Results for Data Fusion and Source Reliability</a>, SIGMOD 2017</li>
<li> Li et al. <a href="http://lunadong.com/publication/webfusion_vldb.pdf">Truth Finding on the Deep Web: Is the Problem Solved?</a>, VLDB 2013 <b>(Optional)</b></li>
<li> Dong et al. <a href="http://lunadong.com/publication/dependence_vldb.pdf">Integrating conflicting data: the role of source dependence</a>, VLDB 2009 <b>(Optional)</b></li>
</ul>
</td>
<td></td>
</tr>
<tr>
<td>13</td>
<td>3/6</td>
<td>Data Wrangling</td>
<td class="side-borders">
<!-- Lecture -->
<a href=""><a href="lectures/Lecture_12_DW.pptx">Lecture 12</a>
</td>
<td class="side-borders">
<ul>
<li> Kandle et al. <a href="http://ilpubs.stanford.edu:8090/1005/1/wranglerPaper.pdf">Wrangler: Interactive Visual Specification of Data Transformation Scripts</a>, CHI 2011</li>
<li> Raza and Gulwani <a href="http://www.aaai.org/ocs/index.php/AAAI/AAAI17/paper/download/15034/13857">Automated Data Extraction Using Predictive Program Synthesis.</a>, AAAI 2017 <b>(Optional)</b></li>
<li> Jin et al. <a href="https://web.eecs.umich.edu/~mrander/pubs/foofah-full.pdf">Foofah: Transforming Data By Example</a>, SIGMOD 2017 <b>(Optional)</b></li>
</ul>
</td>
<td></td>
</tr>
<tr>
<td>14</td>
<td>3/8</td>
<td>Data Cleaning: Error Detection</td>
<td class="side-borders">
<!-- Lecture -->
<a href=""><a href="lectures/Lecture_13_ED.pptx">Lecture 13</a>
</td>
<td class="side-borders">
<ul>
<li> Abedjan et al. <a href="http://www.vldb.org/pvldb/vol9/p993-abedjan.pdf">Detecting Data Errors:
Where are we and what needs to be done?</a>, VLDB 2016</li>
<li> Wu and Madden <a href="http://www.vldb.org/pvldb/vol6/p553-wu.pdf">Scorpion: Explaining away outliers in aggregate queries</a>, VLDB 2013 <b>(Optional)</b></li>
<li> Chu et al. <a href="https://cs.uwaterloo.ca/~ilyas/papers/XuICDE2013.pdf">Holistic data cleaning: Putting violations into context.</a>, ICDE 2013 <b>(Optional)</b></li>
</ul>
</td>
<td></td>
</tr>
<tr>
<td>15</td>
<td>3/13</td>
<td>Data Cleaning: Error Repairing</td>
<td class="side-borders">
<!-- Lecture -->
<a href=""><a href="lectures/Lecture_14_DR.pptx">Lecture 14</a>
</td>
<td class="side-borders">
<ul>
<li> Rekatsinas et al. <a href="http://www.vldb.org/pvldb/vol10/p1190-rekatsinas.pdf">HoloClean: Holistic Data Repairs with Probabilistic Inference</a>, VLDB 2017</li>
<li> Hellesrstein <a href="http://db.cs.berkeley.edu/jmh/papers/cleaning-unece.pdf">Quantitative Data Cleaning for Large Databases</a>, 2008 <b>(Optional)</b></li>
<li> Fagin et al.<a href="papers/p3-fagin.pdf">Dichotomies in the Complexity of Preferred Repairs</a>, PODS 2015 <b>(Optional)</b></li>
</ul>
</td>
<td></td>
</tr>
<tr>
<td>16</td>
<td>3/15</td>
<td>Data Cleaning and Machine Learning</td>
<td class="side-borders">
<!-- Lecture -->
<a href=""><a href="lectures/Lecture_15_CML.pptx">Lecture 15</a>
</td>
<td class="side-borders">
<ul>
<li> Krishnan et al. <a href="http://www.vldb.org/pvldb/vol9/p948-krishnan.pdf">http://www.vldb.org/pvldb/vol9/p948-krishnan.pdf</a>, 2016</li>
<li> Krishnan et al. <a href="https://arxiv.org/abs/1711.01299">BoostClean: Automated Error Detection and Repair for Machine Learning</a>, 2017<b>(Optional)</b></li>
<li> Dolatshah et al.<a href="http://www.cs.sfu.ca/~jnwang/papers/TR-TARS.pdf">Cleaning Crowdsourced Labels Using Oracles For Supervised Learning</a>, 2018 <b>(Optional)</b></li>
</ul>
</td>
<td></td>
</tr>
<tr>
<td>17</td>
<td>3/20</td>
<td>Management of Data under Uncertainty</td>
<td class="side-borders">
<!-- Lecture -->
<a href=""><a href="lectures/Lecture_16_PDB.pptx">Lecture 16</a>
</td>
<td class="side-borders">
<ul>
<li> Dalvi et al. <a href="https://www-cs.stanford.edu/~chrismre/papers/cacm-paper-full.pdf">Probabilistic Databases: Diamonds in the Dirt</a>, 2009</li>
<li> Dalvi et al. <a href="http://www.vldb.org/conf/2004/RS22P1.PDF">Efficient Query Evaluation on Probabilistic Databases</a>, 2004<b>(Optional)</b></li>
</ul>
</td>
<td></td>
</tr>
<tr>
<td>18</td>
<td>3/22</td>
<td>Summary of Data Preparation</td>
<td class="side-borders">
<!-- Lecture -->
<a href="lectures/Lecture_17_recap.pptx">Lecture 17</a>
</td>
<td class="side-borders"> No Readings.</td>
<td></td>
</tr>
<tr>
<td align="center" colspan="6" class="small-section-header">RBMS and ML Integration</td>
</tr>
<tr>
<td>19</td>
<td>4/3</td>
<td>Relational vs Linear Algebra: One semi-ring to rule them all!</td>
<td class="side-borders">
<!-- Lecture -->
<a href="lectures/Lecture_18_SR.pptx">Lecture 18</a>
</td>
<td class="side-borders">
<ul>
<li> Joglekar et al. <a href="hhttps://arxiv.org/pdf/1508.07532.pdf">Aggregations over Generalized Hypertree Decompositions</a>, PODS 2016</li>
<li> Aberger et al. <a href="http://www.aberger.site/levelheaded.pdf">LevelHeaded: A Unified Engine for Business Intelligence and Linear Algebra Querying
</a>, ICDE 2018</li>
<li> Jananthan et al. <a href="https://arxiv.org/pdf/1712.00802.pdf">Polystore Mathematics of Relational Algebra</a>, 2017<b>(Optional)</b></li>
<li> Hutchison et al. <a href="https://arxiv.org/pdf/1703.07342.pdf">LaraDB: A Minimalist Kernel for Linear and Relational Algebra Computation
</a>, BeyondMR 2017<b>(Optional)</b></li>
</ul>
</td>
<td><b>Intermediate Report due!</b></td>
</tr>
<tr>
<td>20</td>
<td>4/5</td>
<td>ML and Data Systems: SQL and UDFs</td>
<td class="side-borders">
<!-- Lecture -->
<a href="lectures/Lecture_19_MAD.pptx">Lecture 19</a>
</td>
<td class="side-borders">
<ul>
<li> Luo et al. <a href="https://pdfs.semanticscholar.org/61e2/f35b4ef5d569aa3084e3118ea16df200965f.pdf">Scalable Linear Algebra on a Relational Database System</a>, 2017</li>
<li> Hellerstein et al. <a href="https://arxiv.org/pdf/1208.4165.pdf">The MADlib Analytics Library or MAD Skills, the SQL</a>, 2012</li>
</ul>
</td>
<td></td>
</tr>
<tr>
<td>21</td>
<td>4/10</td>
<td>ML and Data Systems: Statistical Relational Learning Engines</td>
<td class="side-borders">
<!-- Lecture -->
<a href="lectures/Lecture_20_MRF.pptx">Lecture 20</a>
</td>
<td class="side-borders">
<ul>
<li> Zhang and Re <a href="http://www.vldb.org/pvldb/vol7/p1283-zhang.pdf">DimmWitted: A Study of Main-Memory Statistical Analytics</a>, 2014</li>
<li> Niu et al. <a href="https://www.cs.stanford.edu/people/chrismre/papers/tuffy-vldb11.pdf">Tuffy: Scaling up Statistical Inference in Markov Logic Networks using an RDBMS</a><b> (Optional)</b>, 2011</li>
<li> Tran et al. <a href="https://arxiv.org/pdf/1610.09787.pdf">Edward: A library for probabilistic modeling, inference, and criticism</a><b> (Optional)</b>, 2017</li>
</ul>
</td>
<td></td>
</tr>
<tr>
<td>22</td>
<td>4/12</td>
<td>DB-inspired ML Systems: From Linear Algebra to Execulation Plans and Rewrites</td>
<td class="side-borders">
<!-- Lecture -->
<a href="lectures/Lecture_21_OPT.pptx">Lecture 21</a>
</td>
<td class="side-borders">
<ul><li> Boehm et al. <a href=" https://pdfs.semanticscholar.org/0a9d/a717999ba687bcbef86cbc996f4b7334f990.pdf">SystemML's Optimizer:
Plan Generation for Large-Scale Machine Learning Programs</a>, 2014</li>
</ul>
</td>
<td></td>
</tr>
<tr>
<td>23</td>
<td>4/17</td>
<td>DB-inspired ML Systems: Compress, Scan, Index</td>
<td class="side-borders">
<!-- Lecture -->
<a href="lectures/Lecture_22_COMP.pptx">Lecture 22</a>
</td>
<td class="side-borders">
<ul>
<li> Elgohary et al.<a href="https://researcher.watson.ibm.com/researcher/files/ibm-Matthias.Boehm1/SIGMODRec2017.pdf">Scaling Machine Learning via Compressed Linear Algebra</a>, 2017</li>
<li> Boehm et al. <a href="https://arxiv.org/pdf/1801.00829.pdf">On Optimizing Operator Fusion Plans
for Large-Scale Machine Learning in SystemML</a><b> (Optional)</b>, 2011</li>
</ul>
</td>
<td></td>
</tr>
<tr>
<td>24</td>
<td>4/19</td>
<td>ML serving: Feature engineering</td>
<td class="side-borders">
<a href="lectures/Lecture_23_FS.pptx">Lecture 23</a>
</td>
<td class="side-borders">
<ul>
<li>Zhang et al. <a href="https://cs.stanford.edu/people/chrismre/papers/mod539-zhang.pdf">Materialization optimizations for feature selection workloads</a>, SIGMOD 2014</li>
<li>Nargesian et al. <a href="https://www.ijcai.org/proceedings/2017/0352.pdf">Learning Feature Engineering for Classification</a>, IJCAI 2017 <b>(Optional)</b></li>
</ul>
</td>
<td></td>
</tr>
<tr>
<td>25</td>
<td>4/24</td>
<td>ML serving: Model hubs; repositories for your models</td>
<td class="side-borders">
<!-- Lecture -->
<a href="lectures/Lecture_24_MH.pptx">Lecture 24</a>
</td>
<td class="side-borders">
<ul>
<li>Miao et al. <a href="https://arxiv.org/abs/1611.06224">ModelHub: Towards Unified Data and Lifecycle Management for Deep Learning</a>, 2016</li>
</td>
<td></td>
</tr>
<tr>
<td>26</td>
<td>4/26</td>
<td>Project presentations 1</td>
<td class="side-borders">
<!-- Lecture -->
</td>
<td class="side-borders"></td>
<td></td>
</tr>
<tr>
<td>27</td>
<td>5/1</td>
<td>Project presentations 2</td>
<td class="side-borders">
<!-- Lecture -->
</td>
<td class="side-borders"></td>
<td></td>
</tr>
<tr>
<td>28</td>
<td>5/3</td>
<td>Poster presentations</td>
<td class="side-borders">
<!-- Lecture -->
</td>
<td class="side-borders"></td>
<td></td>
</tr>
</tbody>
</table>
</p>
</div>
</div>
<div class="panel panel-default">
<div class="panel-heading">Grading</div>
<div class="panel-body">
<a name="grading"></a>
<table class="table">
<tr><td>Questions and Comments</td><td>20%</td>
<tr><td>Responses and Discussion</td><td>20%</td>
<tr><td>Project Proposal</td><td>10%</td>
<tr><td>Project Intermediate Report</td><td>10%</td>
<tr><td>Project Final Report</td><td>30%</td>
<tr><td>Project Presentation and Poster</td><td>10%</td>
</table>
</div>
</div>
<div class="panel panel-default">
<div class="panel-heading">Office Hours</div>
<div class="panel-body">
<a name="office-hours"></a>
<p>Theo: by appointment @ Room CS4361</p>
</div>
</div>
<div class="panel panel-default">
<div class="panel-heading">Late Policy and Deliverables</div>
<div class="panel-body">
<a name="late-policy"></a>
There will be <b>no late dates for the project deliverables</b>. However, you have the option to <b>skip up to three QCs</b>. Additional extensions may be granted in the case of a severe medical or family emergency.
</div>
</div>
<div class="panel panel-default">
<div class="panel-heading">Credit</div>
<div class="panel-body">
The template of this website was created by HazyReseach@Stanford.
</div>
</div>
</div>
</body>
</html>