-
Notifications
You must be signed in to change notification settings - Fork 6
Expand file tree
/
Copy pathindex.xhtml
More file actions
311 lines (262 loc) · 19.7 KB
/
index.xhtml
File metadata and controls
311 lines (262 loc) · 19.7 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
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.1//EN" "http://www.w3.org/TR/xhtml11/DTD/xhtml11.dtd">
<html xmlns="http://www.w3.org/1999/xhtml" lang="en-gb" xml:lang="en-gb">
<head>
<!--nomodify-->
<meta name="style" content="cs" />
<title>University of Bristol - Computer Science Department - COMSM0045 - Applied Deep Learning</title>
<link rel="shortcut icon" type="image/ico" href="favicon.png" />
<style>
table { border-spacing: 2px; border-collapse: separate; }
td { margin: 1px; padding: 5px; }
table.blank {vertical-align: top; text-align: left; background: #f1f1f1;}
td.blank {vertical-align: top; text-align: center; background: #cccccc;}
tr.blank {vertical-align: top; text-align: left; background: #cccccc;}
td.mike {vertical-align: top; text-align: left; background: #FFE5B4;}
td.tilo {vertical-align: top; text-align: left; background: #ccffc1;}
td.assess {vertical-align: top; text-align: left; background: #ffaaaa;}
td.labs {vertical-align: top; text-align: left; background: #bbbbee;}
</style>
</head>
<body>
<div id="wrap">
<a href="https://www.bris.ac.uk/unit-programme-catalogue/UnitDetails.jsa?unitCode=COMSM0045" target="_blank">UNIT INFO</a>
<h1>COMSM0045 - Applied Deep Learning</h1>
<img src="comsm0018.jpg" alt="ADL Banner" width="900"></img>
<link rel="stylesheet" href="simple.css" />
<a id="info"></a>
<hr/>
<h2>Unit Information</h2>
<p>Welcome to COMSM0045. The unit introduces the students to deep
architectures for learning linear and non-linear transformations of big
data towards tasks such as classification and regression. The unit paves
the path from understanding the fundamentals of convolutional and
recurrent neural networks through to training and optimisation as well as
evaluation of learnt outcomes. The unit's approach is hands-on, focusing
on the 'how-to' while covering the basic theoretical foundations. For
further general information, see
<a href="https://www.bris.ac.uk/unit-programme-catalogue/UnitDetails.jsa?unitCode=COMSM0045" target="_blank">
the syllabus for the unit</a>.
</p>
<p><b>UPDATE - 22/09/2025 Welcome to the 25/26 cohort!</b></p>
<p><b> PLEASE NOTE: lecture content will be updated, slides below are placeholder and may change until the lecture</b></p>
If you have any questions, head to the <a href = "https://teams.microsoft.com/l/channel/19%3ALWXnse6nfuQqi6JCGIJkA_LL9E4IATzuWivU5VQQqNo1%40thread.tacv2/General?groupId=c1bcad56-70dc-4e9b-ae72-1aa0f583695d&tenantId=b2e47f30-cd7d-4a4e-a5da-b18cf1a4151b">unit teams</a>.
<hr/>
<h2> Staff</h2>
<table>
<tr><td class="mike"><a href="http://mwray.github.io" target="_blank">Michael Wray (MW)</a></td><td><b>Unit Director</b></td></tr>
<tr><td class="tilo"><a href="http://www.cs.bris.ac.uk/~burghard" target="_blank">Tilo Burghardt (TB)</a></td><td></td></tr>
</table>
<hr/>
<h2> Teaching Assistants</h2>
<p>Omar Emara (OE), Prajwal Gatti (PG), Rhodri Guerrier (RG), Sam Pollard (SP), Saptarshi Sinha (SS), Siddhant Bansal (SB), Yini Li (YL)</p>
<a id="materials"></a>
<hr/>
<h2>Unit Materials</h2>
<table class="blank">
<tr class="blank">
<td><i>Wks</i></td> <td><i>Monday 16:00-18:00</i></td> <td><i>Tuesday 15:00-18:00</i></td> <td><i>Labs</i></td>
</tr>
<tr>
<td class="blank">1</td>
<td class="tilo">
<b>22/09/2025 - 16:00 - Queens BLDG 1.07</b><br/>
Wk1 - LECTURE 1<br/>
<b>INTRODUCTION TO THE UNIT</b><br/>
<a href="https://uob-my.sharepoint.com/:p:/g/personal/mw1760_bristol_ac_uk/EafrmxUTdvtIkpTlaFc8CKYBmxfw3dzP7-4M8jThiSJbYw?e=bhwubg" target="_blank">intro slides</a> <br/>
<a href="https://mediasite.bris.ac.uk/Mediasite/Play/7fa0b35cb09b4d4d8b03f16d7a177cbb1d" target="_blank">recording</a>
<hr/>
<b>BASICS OF ARTIFICIAL NEURAL NETWORKS</b><br/>(Queens Building 1.07, in-person)<br/>
(Introduction, Neural Networks, Perceptron, Cost Functions, Gradient Descent, Delta Rule, Deep Networks)<br/>
<a href="https://www.ole.bris.ac.uk/bbcswebdav/courses/COMSM0045_2022_TB-1/content/COMSM0045_01.pdf" target="_blank">PDF Slides</a>, <a href="https://mediasite.bris.ac.uk/Mediasite/Play/fbd084a2bca141c5a663fa076c54f4f91d" target="_blank">Recording</a>
</td>
<td class="tilo">
<b>23/09/2025 - 15:00 - Queens BLDG 1.07</b><br/>
Wk1 - LECTURE 2<br/>
<b>TOWARDS TRAINING DEEP FORWARD NETWORKS</b><br/>(MVB 1.15, in-person + lecture recap)<br/>
(Network Representation, Computational Graphs, Reverse Auto-Differentiation)<br/>
<a href="https://www.ole.bris.ac.uk/bbcswebdav/courses/COMSM0045_2022_TB-1/content/COMSM0045_02.pdf" target="_blank">PDF Slides</a>,
<a href="https://mediasite.bris.ac.uk/Mediasite/Play/4db6e3234d314a1aa2935d71c02a0e071d" target="_blank">Extra Recap Recording Lecture 2 Refresher (first part of video)</a>
</td>
<td><b>GETTING STARTED:</b><br/>
<hr/><a href="#bc4" target="_blank">Register Individually on BlueCrystal4<br/>(details see below)</a><hr/> <b>RECAP WORKSHEETS:</b><br/>-<a href="https://github.com/COMSM0045-Applied-Deep-Learning/labsheets">Lab0 - Python (Homework)</a>
</td> </tr> <tr>
<td class="blank">2</td>
<td class="tilo">
<b>29/09/2025 - 16:00 - Queens BLDG 1.07</b><br/>
Wk2 - LECTURE 3<br/>
<b>BACKPROPAGATION ALGORITHM</b><br/>(Queens Building 1.07, in-person + recorded lecture)<br/>
(The Backpropagation Algorithm in Full Detail, Activation Functions)<br/>
<a href="https://www.ole.bris.ac.uk/bbcswebdav/courses/COMSM0045_2022_TB-1/content/COMSM0045_03.pdf" target="_blank">PDF Slides</a><br/>
<a href="https://mediasite.bris.ac.uk/Mediasite/Play/4db6e3234d314a1aa2935d71c02a0e071d" target="_blank">Extra Recap Recording (second part of video)</a><hr/>
Wk2 - LECTURE 4<br/>
<b>OPTIMISATION TECHNIQUES</b><br/>(Queens Building 1.07, in-person + recorded lecture)<br/>
(Stochastic Gradient Descent, Nesterov Momentum, RMSProp, Newton's Method, AdaGrad, Adam, Saddle Points)<br/>
<a href="https://www.ole.bris.ac.uk/bbcswebdav/courses/COMSM0045_2022_TB-1/content/COMSM0045_04.pdf" target="_blank">PDF Slides</a><br/>
<a href="https://mediasite.bris.ac.uk/Mediasite/Play/aa25459da02e430a8c99e3cfd5082a371d" target="_blank">Extra Recap Recording</a><hr/>
</td>
<td class="mike">
<b>30/09/2025 - 15:00 - MVB 1.15 - </b><br/>PRACTICAL 1<br/>
Your first fully connected layer<br/>
gradient descent<br/>
stochastic gradient descent <br/>
<a href="https://uob-my.sharepoint.com/:p:/g/personal/mw1760_bristol_ac_uk/EcDZBeOxRcFNvv84jdZSkt0BukAao47UvVJzYin373DboQ?e=IOXenS">Slides</a><br/>
<a href="https://uob-my.sharepoint.com/:v:/g/personal/mw1760_bristol_ac_uk/EaSWGEqdkGpJu8u3Qtn6IxkBaUUjE5Wa6g8rxvd7bqq3vQ?nav=eyJyZWZlcnJhbEluZm8iOnsicmVmZXJyYWxBcHAiOiJTdHJlYW1XZWJBcHAiLCJyZWZlcnJhbFZpZXciOiJTaGFyZURpYWxvZy1MaW5rIiwicmVmZXJyYWxBcHBQbGF0Zm9ybSI6IldlYiIsInJlZmVycmFsTW9kZSI6InZpZXcifX0%3D&e=dSARWX">Recording</a>
</td>
<td class="labs">30/09/2025, (MVB 1.15) - 3hrs<br/>-<a href="https://github.com/COMSM0045-Applied-Deep-Learning/labsheets/blob/master/lab-1-dnns/bc4-setup.ipynb">BC4 Setup</a>
<hr/>
<a href="https://github.com/COMSM0045-Applied-Deep-Learning/labsheets"><b>Lab 1</b> - Training your first Deep Neural Network</a>
</td>
</tr>
<tr>
<td class="blank" rowspan="2">3</td>
<td class="tilo">
<b>06/10/2025 - 16:00 - Queens BLDG 1.07</b><br/>
Wk3 - LECTURE 5<br/>
<b>COST FUNCTIONS, REGULARISATION AND DEPTH</b><br/>
(Queens Building 1.07, in-person + recorded lecture)<br/>
(SoftMax, Cross Entropy, L1 and L2 Regularisation, DropOut, DropConnect, Depth Considerations)<br/>
<a href="https://www.ole.bris.ac.uk/bbcswebdav/courses/COMSM0045_2022_TB-1/content/COMSM0045_05.pdf" target="_blank">PDF Slides</a><br/>
<a href="https://mediasite.bris.ac.uk/Mediasite/Play/9fd25a71068443d18c1689759d3e79f31d" target="_blank">Extra Recap Recording</a>
</td>
<td class="mike" rowspan="2">
<b>07/10/2025 - 15:00 - MVB 1.15 - </b><br/>PRACTICAL 2<br/>
Your first convolutional connected layer<br/>
<a href="https://uob-my.sharepoint.com/:p:/g/personal/mw1760_bristol_ac_uk/ERsmes4RuE9OgfRvc6h3c8EBJOgGSdhGdwdAWd1K8r8j1A?e=uo9fqc">Slides</a>
<br/>
<a target="_blank" href="https://uob-my.sharepoint.com/:v:/g/personal/mw1760_bristol_ac_uk/ERsXMz5XynlMshVrwMR31WoBBowHaYzlfuDV7yojpHxXKg?nav=eyJyZWZlcnJhbEluZm8iOnsicmVmZXJyYWxBcHAiOiJTdHJlYW1XZWJBcHAiLCJyZWZlcnJhbFZpZXciOiJTaGFyZURpYWxvZy1MaW5rIiwicmVmZXJyYWxBcHBQbGF0Zm9ybSI6IldlYiIsInJlZmVycmFsTW9kZSI6InZpZXcifX0%3D&e=lnCe2B">Recording</a>
</td>
<td class="labs" rowspan="2">07/10/2025, (MVB 1.15) - 3hrs<br/> <a href="https://github.com/COMSM0045-Applied-Deep-Learning/labsheets"><b>Lab 2</b> - Your First Convolutional Connected Network</a></td>
</tr>
<tr>
<td class="mike">
Wk3 - LECTURE 6<br/><b>CONVOLUTIONAL NEURAL NETWORKS</b><br/>(Queens Building 1.07, in-person + recorded lecture)<br/>
(sharing parameters, conv layers, pooling, CNN architectures)<br/>
<a target="_blank" href="https://uob-my.sharepoint.com/:p:/g/personal/mw1760_bristol_ac_uk/ERoeviEnV5pCoRks91CKfqwB171GwESuQXMcVDLmmAcd8A?e=5LKirk">Slides</a>
<br/>
<a target="_blank" href="https://mediasite.bris.ac.uk/Mediasite/Play/56921b3aeae4422d8c183334f72d9f011d">Recording</a>
</td>
</tr>
<tr>
<td class="blank">4</td> <td class="mike">
<b>13/10/2025 - 16:00 - Queens BLDG 1.07</b><br/>
Wk4 - LECTURE 7 <br/>
<b>RECURRENT and RELATIONAL NEURAL NETWORKS</b><br/>(Queens Building 1.07, in-person + Recorded) <br/>
(RNN, encoder-decoder, Transformers)<br/>
<a href="https://uob-my.sharepoint.com/:p:/g/personal/mw1760_bristol_ac_uk/EZBKWDIOkU9LjOVzWCcJLs4BVUXNGgs99bV0bUPUVY8WiA?e=1S6a48">Slides</a>
<br/>
<a href="https://mediasite.bris.ac.uk/Mediasite/Play/a2b82230f50b4f6d84dfd3c14a4f0cdd1d">Recording Pt. 1</a>
<a href="https://mediasite.bris.ac.uk/Mediasite/Play/4c22368682e44998a246f3460bd6f2061d">Recording Pt. 2</a>
</td>
<td class="mike">
<b>14/10/2025 - 15:00 - MVB 1.15 - </b>
<br/>PRACTICAL 3<br/>
Error rate monitoring (training/validation/testing)<br/>
Batch-based training<br/>
Learning rate<br/>
Weight Freezing<br/>
Batch normalisation<br/>
Parameter intialisation<br/>
<a href="https://uob-my.sharepoint.com/:p:/g/personal/mw1760_bristol_ac_uk/EVPEDTlo1qpMrPhWUkdURUIB9CDz52IrhdsOWY5vAgl8wg?e=DWE1W0">Slides</a>
<br/>
<a href="https://uob-my.sharepoint.com/:v:/g/personal/mw1760_bristol_ac_uk/ER4fL-5iovlKtZZSY30_zikBH89CsP5B4sjSL0EV5O77kw?nav=eyJyZWZlcnJhbEluZm8iOnsicmVmZXJyYWxBcHAiOiJPbmVEcml2ZUZvckJ1c2luZXNzIiwicmVmZXJyYWxBcHBQbGF0Zm9ybSI6IldlYiIsInJlZmVycmFsTW9kZSI6InZpZXciLCJyZWZlcnJhbFZpZXciOiJNeUZpbGVzTGlua0NvcHkifX0&e=m4m1bG">Video Recording</a>
</td>
<td class="labs">14/10/2025, (MVB 1.15) - 3hrs<br/><br/><a href="https://github.com/COMSM0045-Applied-Deep-Learning/labsheets"><b>Lab 3</b> - Hyperparameters</a></td> </tr>
<tr> <td class="blank">5</td>
<td class="mike">
<b>20/10/2025 - 16:00 - Queens BLDG 1.07</b><br/>
Wk5 - LECTURE 8 <br/>
<b>GENERATIVE MODELS</b><br/>
(Queens Building 1.07, in-person + Recorded) <br/>
(Autoregressive models)<br/>
<a href="https://uob-my.sharepoint.com/:p:/g/personal/mw1760_bristol_ac_uk/EbDVd6dVd8tJvU7epyQumSkBlK1Hj81gN4P0Q_0jo5dHrg?e=ZXtInF">Slides</a>
<br/>
<a href="https://mediasite.bris.ac.uk/Mediasite/Play/4c22368682e44998a246f3460bd6f2061d">Recording</a>
</td>
<td class="mike">
<b>21/10/2025 - 15:00 - MVB 1.15 - </b><br/>PRACTICAL 4<br/> Data Augmentation<br/>Debugging strategies<br/>
Dropout
<br/>
<a href="https://uob-my.sharepoint.com/:p:/g/personal/mw1760_bristol_ac_uk/ESe18HydSJNKir90YXGZAGUBPfXOBSXFqDT3cs9LMWU7mw?e=cHpjfW">Slides</a>
<br/>
<a href="https://uob-my.sharepoint.com/:v:/g/personal/mw1760_bristol_ac_uk/EXzwPgmKfw1DgZ5G_fmI87UB_k-lKxlIiWNvXHE6AwEwKQ?nav=eyJyZWZlcnJhbEluZm8iOnsicmVmZXJyYWxBcHAiOiJPbmVEcml2ZUZvckJ1c2luZXNzIiwicmVmZXJyYWxBcHBQbGF0Zm9ybSI6IldlYiIsInJlZmVycmFsTW9kZSI6InZpZXciLCJyZWZlcnJhbFZpZXciOiJNeUZpbGVzTGlua0NvcHkifX0&e=4bbmNM">Video Recording</a><br/>
</td>
<td class="labs">21/10/2025, (MVB 1.15) - 3hrs<br/><br/><a href="https://github.com/COMSM0045-Applied-Deep-Learning/labsheets"><b>Lab 4</b> - Data Augmentation</a></td></tr>
<tr> <td class="blank">6</td> <td class="blank" colspan="3">READING WEEK - Mid Term for MAJOR unit students 30/10/2025 - MVB - 1.07 - 13:00-14:00 </td></tr>
<tr>
<td class="blank">7</td>
<td class="blank">-</td>
<td class="mike">
<b>04/11/2025 - 15:00 - MVB 1.15 - </b><br/>PRACTICAL 5<br/> Transformers<br/>
<a href="https://uob-my.sharepoint.com/:p:/g/personal/mw1760_bristol_ac_uk/EQnpCL_A6M1En_409FdFeFkB2oHdV1MkWQwT50UhoETJ7A?e=m8SCI3">Slides</a>
<br/>
</td>
<td class="labs">04/11/2025, (MVB 1.15) - 3hrs<br/><a href="https://github.com/COMSM0045-Applied-Deep-Learning/labsheets"><b>Lab 5</b> - Transformers</a></td>
</tr>
<tr>
<td class="blank">8</td>
<td class="blank">-</td>
<td class="mike"><b>11/11/2025 - 15:00 - MVB 1.15 - </b><br/>
Continuation Lab<br/> </td>
<td class="labs">11/11/2025, (MVB 1.15) - 3hrs<br/><br/> </td>
</tr>
<tr> <td class="blank">9</td> <td class="blank">-</td> <td class="mike"><b>18/11/2025, 15:00 [2 hours], (MVB 1.15) - </b><br/>CW Support Session</td>
<td class="blank">-</td></tr>
<tr>
<td class="blank">10</td> <td class="blank">-</td> <td class="mike"><b>25/11/2025, 15:00 [2 hours], (MVB 1.15) - </b><br/>CW Support Session</td>
<td class="blank">-</td>
</tr>
<tr>
<td class="blank">11</td> <td class="blank">-</td> <td class="mike"><b>02/12/2025, 15:00 [2 hours], (MVB 1.15) - </b><br/>CW Support Session</td>
<td class="blank">-</td>
</tr>
<tr> <td class="blank">12</td> <td class="blank">-</td> <td class="mike"><b>09/12/2025, Queens BLDG 1.07 - </b><br/>Exam Support Session</td><td class="blank">-</td></tr>
<tr> <td class="blank">13</td> <td class="blank" colspan="3">DECEMBER EXAMS - Final for MINOR unit students </td></tr>
</table>
<hr/>
<h2>Assessment Details</h2>
<ul>
<li><b>Coursework (for Major option): </b> You are requested to re-implement work to be specified, and provide your code as well as a final report. Coursework completed in groups (up to 3)</li>
<li><b>Mid-term test (for Major option): </b> The theoretical components of the unit up to week 5 are examinable. No code would be written in the exam, but code can be provided to answer questions. Calculators are allowed.</li>
<li><b>Exam (for Minor option): </b> Both the theoretical and practical components of the unit are examinable. No code would be written in the exam, but code can be provided to answer questions. Calculators are allowed.</li>
</ul>
<p> </p>
<hr/>
<h2>Assessment Details - Coursework</h2>
<p>The coursework will be released during TB1 </p>
<!--<p>Note that the deadline is the 29th November at 13:00 </p>
<p><a href="https://forms.office.com/e/3DYE6Qm6Z4">Form for Sign-Ups</a>. Even if you plan to work solo, please fill in this form. </p>
<p> If you need help, ask on the unit teams and/or come to the weekly CW support sessions. </p>-->
<hr/>
<h2>Assessment Details - Exam</h2>
More details regarding the exam and in-class test coming soon.
You can find previous papers <a href="https://uob-my.sharepoint.com/:f:/g/personal/mw1760_bristol_ac_uk/EmldHGtc0MBLoUErTEubMj0BFDlCQzibjgjMNW40uAcMug?e=U2vkpb">here</a>, but please note that these were from when the unit only contained one 2 hour exam.
<!--<p> We have now released the exam preparation materials on the unit webpage, you can also find these in the folder <a href="https://uob-my.sharepoint.com/:f:/g/personal/mw1760_bristol_ac_uk/EmldHGtc0MBLoUErTEubMj0BFDlCQzibjgjMNW40uAcMug?e=U2vkpb">here</a>.</p>
<p> This includes a list of exam topics, 3 past papers (1 with answers). </p>-->
<p> Please note that you cannot take notes into the exam (it is closed book), but calculators are permitted.</p>
<hr/>
<h2>Github</h2>
<p>All technical resources will be posted on the
<a href="https://github.com/COMSM0045-Applied-Deep-Learning" target="_blank">COMSM0045 ADL Github organisation</a>. If you find any issues, please kindly raise an issue in the respective repository.
</p>
<hr/>
<h2>Textbook</h2>
<p>Recommended Reading:<a href="https://udlbook.github.io/udlbook/" target="_blank">Simon J.D (2023). Prince. Understanding Deep Learning, MIT Press</a><br/></p>
You can also check out, written pre transformers, the older course book which we still recommend: <a href="http://www.deeplearningbook.org/" target="_blank">Goodfellow et al (2016). Deep Learning. MIT Press</a>
<hr/>
<!--<h2><a id="bc4">Blue Crystal 4 Registration [only applicable for Bristol undergraduate students with corresponding email]</a></h2>
<p>All students must apply online to register an account on BC4 for this
unit. This also applies to students who already have accounts on BC4 for other units (e.g. HPC), in this case you must
register again using the instructions below.</p>
<ol>
<li>Click on: <a href="https://www.acrc.bris.ac.uk/login-area/apply.cgi" target="_blank">https://www.acrc.bris.ac.uk/login-area/apply.cgi</a></li>
<li>Enter your personal details</li>
<li>Choose: "Join an existing project"</li>
<li>Enter project code: COMS033444</li>
<li>Keep Preferred log-in shell as bash</li>
<li>In the comments box please enter the following:</li>
"I am on the taught course Applied Deep Learning (COMSM0045), unit director Michael Wray, and will need access to BC4"
</ol>
<p>Note that it takes up to 48 hours to enable your account on BC4.</p>-->
</div>
</body>
</html>