Loaded: 4%Progress bar: 4.

75%PlayRewind 5s1xPlayback
rate0.5x0.75x1x1.25x1.5x1.75x2xForward 5s0:40 / 14:154:44 / 44:44Add
noteTranscriptCaptionsOffEnglish [CC]Czech [Auto]Dutch [Auto]Finnish [Auto]French
[Auto]German [Auto]Indonesian [Auto]Italian [Auto]Japanese [Auto]Korean
[Auto]Polish [Auto]Portuguese [Auto]Swedish [Auto]Spanish [Auto]Simplified Chinese
[Auto]Turkish [Auto]Vietnamese [Auto]Caption
settingsSettings1080p720p576p432p360pAutoAutoplayDownload lectureKeyboard
shortcutsContent informationReport abuseFullscreenBookmark title8 people have
written a note here.Bookmark title17 people have written a note here.Bookmark
title13 people have written a note here.Bookmark title6 people have written a note
here.Bookmark title13 people have written a note here.Bookmark title7 people have
written a note here.Bookmark title12 people have written a note here.29.
Derivatives: product and chain rules31. What about local minima?Course contentAI
AssistantSection 1: Introduction2 / 2 | 17min2 of 2 lectures completed17minSection
2: Download all course materials2 / 2 | 8min2 of 2 lectures completed8minSection 3:
Concepts in deep learning3 / 5 | 1hr 16min3 of 5 lectures completed1hr 16minSection
4: About the Python tutorial0 / 1 | 4min0 of 1 lecture completed4minSection 5:
Math, numpy, PyTorch1 / 19 | 3hr 21min1 of 19 lectures completed3hr 21minSection 6:
Gradient descent10 / 10 | 1hr 57min10 of 10 lectures completed1hr 57minPlay30.
Overview of gradient descent14minPlay31. What about local minima?12minPlay32.
Gradient descent in 1D17minPlay33. CodeChallenge: unfortunate starting
value12minPlay34. Gradient descent in 2D15minPlay35. CodeChallenge: 2D gradient
ascent5minPlay36. Parametric experiments on g.d.19minPlay37. CodeChallenge: fixed
vs. dynamic learning rate16minPlay38. Vanishing and exploding gradients6minPlay39.
Tangent: Notebook revision history2minSection 7: ANNs (Artificial Neural Networks)0
/ 21 | 5hr 12min0 of 21 lectures completed5hr 12minSection 8: Overfitting and
cross-validation0 / 8 | 1hr 48min0 of 8 lectures completed1hr 48minSection 9:
Regularization0 / 12 | 2hr 38min0 of 12 lectures completed2hr 38minSection 10:
Metaparameters (activations, optimizers)0 / 24 | 4hr 52min0 of 24 lectures
completed4hr 52minSection 11: FFNs (Feed-Forward Networks)0 / 12 | 2hr 15min0 of 12
lectures completed2hr 15minSection 12: More on data0 / 11 | 2hr 25min0 of 11
lectures completed2hr 25minSection 13: Measuring model performance0 / 8 | 1hr
20min0 of 8 lectures completed1hr 20minSection 14: FFN milestone projects0 / 6 |
1hr 2min0 of 6 lectures completed1hr 2minSection 15: Weight inits and
investigations0 / 10 | 2hr 19min0 of 10 lectures completed2hr 19minSection 16:
Autoencoders0 / 6 | 1hr 51min0 of 6 lectures completed1hr 51minSection 17: Running
models on a GPU0 / 3 | 32min0 of 3 lectures completed32minSection 18: Convolution
and transformations0 / 12 | 2hr 57min0 of 12 lectures completed2hr 57minSection 19:
Understand and design CNNs0 / 16 | 4hr 10min0 of 16 lectures completed4hr
10minSection 20: CNN milestone projects0 / 5 | 40min0 of 5 lectures
completed40minSection 21: Transfer learning0 / 8 | 1hr 47min0 of 8 lectures
completed1hr 47minSection 22: Style transfer0 / 5 | 58min0 of 5 lectures
completed58minSection 23: Generative adversarial networks0 / 7 | 1hr 25min0 of 7
lectures completed1hr 25minSection 24: RNNs (Recurrent Neural Networks) (and
GRU/LSTM)0 / 9 | 2hr 48min0 of 9 lectures completed2hr 48minSection 25: Ethics of
deep learning0 / 5 | 49min0 of 5 lectures completed49minSection 26: Where to go
from here?0 / 2 | 24min0 of 2 lectures completed24minSection 27: Python intro: Data
types0 / 8 | 1hr 41min0 of 8 lectures completed1hr 41minSection 28: Python intro:
Indexing, slicing0 / 2 | 24min0 of 2 lectures completed24minSection 29: Python
intro: Functions0 / 8 | 1hr 41min0 of 8 lectures completed1hr 41minSection 30:
Python intro: Flow control0 / 10 | 2hr 36min0 of 10 lectures completed2hr
36minSection 31: Python intro: Text and plots0 / 7 | 1hr 42min0 of 7 lectures
completed1hr 42minSection 32: Bonus section0 / 1 | 1min0 of 1 lecture
completed1minExpand panelClose panelOverviewQ&AQuestions and
answersNotesAnnouncementsReviewsLearning toolsDownload ContentSearch course
contentStart a new searchTo find captions, lectures or resourcesMaster deep
learning in PyTorch using an experimental scientific approach, with lots of
examples and practice problems.Rating: 4.8 out of 54.85,111
ratings42,259Students57.5 hoursTotalLast updated April 2025EnglishEnglish [Auto],
Czech [Auto], 15 moreinformation alertSchedule learning timeLearning a little each
day adds up. Research shows that students who make learning a habit are more likely
to reach their goals. Set time aside to learn and get reminders using your learning
scheduler.Get startedDismissBy the numbersSkill level: Beginner LevelStudents:
42259Languages: EnglishCaptions: YesLectures: 265Video: 57.5 total
hoursCertificatesGet Udemy certificate by completing entire courseUdemy
certificateFeaturesAvailable on iOS and AndroidDescriptionDeep learning is
increasingly dominating technology and has major implications for society.From
self-driving cars to medical diagnoses, from face recognition to deep fakes, and
from language translation to music generation, deep learning is spreading like
wildfire throughout all areas of modern technology.But deep learning is not only
about super-fancy, cutting-edge, highly sophisticated applications. Deep learning
is increasingly becoming a standard tool in machine-learning, data science, and
statistics. Deep learning is used by small startups for data mining and dimension
reduction, by governments for detecting tax evasion, and by scientists for
detecting patterns in their research data.Deep learning is now used in most areas
of technology, business, and entertainment. And it's becoming more important every
year.How does deep learning work?Deep learning is built on a really simple
principle: Take a super-simple algorithm (weighted sum and nonlinearity), and
repeat it many many times until the result is an incredibly complex and
sophisticated learned representation of the data.Is it really that simple? mmm OK,
it's actually a tiny bit more complicated than that ;) but that's the core idea,
and everything else -- literally everything else in deep learning -- is just clever
ways of putting together these fundamental building blocks. That doesn't mean the
deep neural networks are trivial to understand: there are important architectural
differences between feedforward networks, convolutional networks, and recurrent
networks.Given the diversity of deep learning model designs, parameters, and
applications, you can only learn deep learning -- I mean, really learn deep
learning, not just have superficial knowledge from a youtube video -- by having an
experienced teacher guide you through the math, implementations, and reasoning. And
of course, you need to have lots of hands-on examples and practice problems to work
through. Deep learning is basically just applied math, and, as everyone knows, math
is not a spectator sport!What is this course all about?Simply put: The purpose of
this course is to provide a deep-dive into deep learning. You will gain flexible,
fundamental, and lasting expertise on deep learning. You will have a deep
understanding of the fundamental concepts in deep learning, so that you will be
able to learn new topics and trends that emerge in the future.Please note: This is
not a course for someone who wants a quick overview of deep learning with a few
solved examples. Instead, this course is designed for people who really want to
understand how and why deep learning works; when and how to select metaparameters
like optimizers, normalizations, and learning rates; how to evaluate the
performance of deep neural network models; and how to modify and adapt existing
models to solve new problems.You can learn everything about deep learning in this
course.In this course, you will learn Theory: Why are deep learning models built
the way they are? Math: What are the formulas and mechanisms of deep learning?
Implementation: How are deep learning models actually constructed in Python (using
the PyTorch library)?Intuition: Why is this or that metaparameter the right choice?
How to interpret the effects of regularization? etc.Python: If you're completely
new to Python, go through the 8+ hour coding tutorial appendix. If you're already a
knowledgeable coder, then you'll still learn some new tricks and code
optimizations.Google-colab: Colab is an amazing online tool for running Python
code, simulations, and heavy computations using Google's cloud services. No need to
install anything on your computer.Unique aspects of this courseClear and
comprehensible explanations of concepts in deep learning, including transfer
learning, generative modeling, convolutional neural networks, feedforward networks,
generative adversarial networks (GAN), and more.Several distinct explanations of
the same ideas, which is a proven technique for learning.Visualizations using
graphs, numbers, and spaces that provide intuition of artificial neural
networks.LOTS of exercises, projects, code-challenges, suggestions for exploring
the code. You learn best by doing it yourself!Active Q&A forum where you can ask
questions, get feedback, and contribute to the community.8+ hour Python tutorial.
That means you don't need to master Python before enrolling in this course.So what
are you waiting for??Watch the course introductory video and free sample videos to
learn more about the contents of this course and about my teaching style. If you
are unsure if this course is right for you and want to learn more, feel free to
contact with me questions before you sign up.I hope to see you soon in the course!
MikeWhat you’ll learnThe theory and math underlying deep learningHow to build
artificial
neural networksArchitectures of feedforward and convolutional networksBuilding
models in PyTorchThe calculus and code of gradient descentFine-tuning deep network
modelsLearn Python from scratch (no prior coding experience necessary)How and why
autoencoders workHow to use transfer learningImproving model performance using
regularizationOptimizing weight initializationsUnderstand image convolution using
predefined and learned kernelsWhether deep learning models are understandable or
mysterious black-boxes!Using GPUs for deep learning (much faster than CPUs!)Are
there any course requirements or prerequisites?Interest in learning about deep
learning!Python/Pytorch skills are taught in the courseA Google account (google-
colab is used as the Python IDE)Who this course is for:Students in a deep learning
courseMachine-learning enthusiastsAnyone interested in mechanisms of AI (artificial
intelligence)Data scientists who want to expand their library of skillsAspiring
data scientistsScientists and researchers interested in deep learningInstructorMike
X CohenEducator and writerI am a full-time educator and writer, and former
professor of neuroscience. I "retired" from that position so I could focus my time
and energy creating high-quality educational material just for you.I have 20 years
of experience teaching programming, data analysis, signal processing, statistics,
linear algebra, and experiment design. I've taught undergraduate students, PhD
candidates, postdoctoral researchers, and full professors. I have taught in
"traditional" university courses, special week-long intensive courses, and Nobel
prize-winning research labs. I have >100 hours of online lectures on neuroscience
data analysis that you can find on my website and youtube channel. And I've written
several technical books about these topics with a few more on the way.I'm not
trying to show off -- I'm trying to convince you that you've come to the right
place to maximize your learning from an instructor who has spent two decades
refining and perfecting his teaching style.Over 300,000 students have watched over
45,000,000 minutes of my courses. Come find out why!I have several free courses
that you can enroll in. Try them out! You got nothing to lose ;)
-------------------------By popular request, here are suggested course progressions
for various educational goals:MATLAB programming: Get Started with MATLAB; Master
MATLAB; Image ProcessingPython programming: Master Python programming by solving
scientific projects; Master Math by Coding in PythonApplied linear algebra:
Complete Linear Algebra; Dimension ReductionSignal processing: Understand the
Fourier Transform; Generate and visualize data; Signal Processing; Neural signal
processingShow moreShow lessCreate a new note at 0:40All lecturesAll
lecturesCurrent lectureSort by most recentSort by most recentSort by oldestClick
the "Create a new note" box, the "+" button, or press "B" to make your first
note.Learning remindersSet up push notifications or calendar events to stay on
track for your learning goals.Add a learning reminderMobile notificationsReceive
learning reminders on your mobile device.Text me a link to download the appPhone
country code+93 (Afghanistan)+358 (Åland Islands)+355 (Albania)+213 (Algeria)+1684
(American Samoa)+376 (Andorra)+244 (Angola)+1264 (Anguilla)+672 (Antarctica)+1268
(Antigua and Barbuda)+54 (Argentina)+374 (Armenia)+297 (Aruba)+61 (Australia)+43
(Austria)+994 (Azerbaijan)+1242 (Bahamas)+973 (Bahrain)+880 (Bangladesh)+1246
(Barbados)+375 (Belarus)+32 (Belgium)+501 (Belize)+229 (Benin)+1441 (Bermuda)+975
(Bhutan)+591 (Bolivia, Plurinational State of bolivia)+387 (Bosnia and Herzegovina)
+267 (Botswana)+47 (Bouvet Island)+55 (Brazil)+246 (British Indian Ocean Territory)
+673 (Brunei Darussalam)+359 (Bulgaria)+226 (Burkina Faso)+257 (Burundi)+855
(Cambodia)+237 (Cameroon)+1 (Canada)+238 (Cape Verde)+ 345 (Cayman Islands)+236
(Central African Republic)+235 (Chad)+56 (Chile)+86 (China)+61 (Christmas Island)
+61 (Cocos (Keeling) Islands)+57 (Colombia)+269 (Comoros)+242 (Congo)+243 (Congo,
The Democratic Republic of the Congo)+682 (Cook Islands)+506 (Costa Rica)+225 (Cote
d'Ivoire)+385 (Croatia)+53 (Cuba)+357 (Cyprus)+420 (Czech Republic)+45 (Denmark)
+253 (Djibouti)+1767 (Dominica)+1849 (Dominican Republic)+593 (Ecuador)+20 (Egypt)
+503 (El Salvador)+240 (Equatorial Guinea)+291 (Eritrea)+372 (Estonia)+251
(Ethiopia)+500 (Falkland Islands (Malvinas))+298 (Faroe Islands)+679 (Fiji)+358
(Finland)+33 (France)+594 (French Guiana)+689 (French Polynesia)+262 (French
Southern Territories)+241 (Gabon)+220 (Gambia)+995 (Georgia)+49 (Germany)+233
(Ghana)+350 (Gibraltar)+30 (Greece)+299 (Greenland)+1473 (Grenada)+590 (Guadeloupe)
+1671 (Guam)+502 (Guatemala)+44 (Guernsey)+224 (Guinea)+245 (Guinea-Bissau)+592
(Guyana)+509 (Haiti)+0 (Heard Island and Mcdonald Islands)+379 (Holy See (Vatican
City State))+504 (Honduras)+852 (Hong Kong)+36 (Hungary)+354 (Iceland)+91 (India)
+62 (Indonesia)+98 (Iran, Islamic Republic of Persian Gulf)+964 (Iraq)+353
(Ireland)+44 (Isle of Man)+972 (Israel)+39 (Italy)+1876 (Jamaica)+81 (Japan)+44
(Jersey)+962 (Jordan)+7 (Kazakhstan)+254 (Kenya)+686 (Kiribati)+850 (Korea,
Democratic People's Republic of Korea)+82 (Korea, Republic of South Korea)+383
(Kosovo)+965 (Kuwait)+996 (Kyrgyzstan)+856 (Laos)+371 (Latvia)+961 (Lebanon)+266
(Lesotho)+231 (Liberia)+218 (Libyan Arab Jamahiriya)+423 (Liechtenstein)+370
(Lithuania)+352 (Luxembourg)+853 (Macao)+389 (Macedonia)+261 (Madagascar)+265
(Malawi)+60 (Malaysia)+960 (Maldives)+223 (Mali)+356 (Malta)+692 (Marshall Islands)
+596 (Martinique)+222 (Mauritania)+230 (Mauritius)+262 (Mayotte)+52 (Mexico)+691
(Micronesia, Federated States of Micronesia)+373 (Moldova)+377 (Monaco)+976
(Mongolia)+382 (Montenegro)+1664 (Montserrat)+212 (Morocco)+258 (Mozambique)+95
(Myanmar)+264 (Namibia)+674 (Nauru)+977 (Nepal)+31 (Netherlands)+599 (Netherlands
Antilles)+687 (New Caledonia)+64 (New Zealand)+505 (Nicaragua)+227 (Niger)+234
(Nigeria)+683 (Niue)+672 (Norfolk Island)+1670 (Northern Mariana Islands)+47
(Norway)+968 (Oman)+92 (Pakistan)+680 (Palau)+970 (Palestinian Territory, Occupied)
+507 (Panama)+675 (Papua New Guinea)+595 (Paraguay)+51 (Peru)+63 (Philippines)+64
(Pitcairn)+48 (Poland)+351 (Portugal)+1939 (Puerto Rico)+974 (Qatar)+40 (Romania)+7
(Russia)+250 (Rwanda)+262 (Reunion)+590 (Saint Barthelemy)+290 (Saint Helena,
Ascension and Tristan Da Cunha)+1869 (Saint Kitts and Nevis)+1758 (Saint Lucia)+590
(Saint Martin)+508 (Saint Pierre and Miquelon)+1784 (Saint Vincent and the
Grenadines)+685 (Samoa)+378 (San Marino)+239 (Sao Tome and Principe)+966 (Saudi
Arabia)+221 (Senegal)+381 (Serbia)+248 (Seychelles)+232 (Sierra Leone)+65
(Singapore)+421 (Slovakia)+386 (Slovenia)+677 (Solomon Islands)+252 (Somalia)+27
(South Africa)+211 (South Sudan)+500 (South Georgia and the South Sandwich Islands)
+34 (Spain)+94 (Sri Lanka)+249 (Sudan)+597 (Suriname)+47 (Svalbard and Jan Mayen)
+268 (Swaziland)+46 (Sweden)+41 (Switzerland)+963 (Syrian Arab Republic)+886
(Taiwan)+992 (Tajikistan)+255 (Tanzania, United Republic of Tanzania)+66 (Thailand)
+670 (Timor-Leste)+228 (Togo)+690 (Tokelau)+676 (Tonga)+1868 (Trinidad and Tobago)
+216 (Tunisia)+90 (Turkey)+993 (Turkmenistan)+1649 (Turks and Caicos Islands)+688
(Tuvalu)+256 (Uganda)+380 (Ukraine)+971 (United Arab Emirates)+44 (United Kingdom)
+1 (United States)+598 (Uruguay)+998 (Uzbekistan)+678 (Vanuatu)+58 (Venezuela,
Bolivarian Republic of Venezuela)+84 (Vietnam)+1284 (Virgin Islands, British)+1340
(Virgin Islands, U.S.)+681 (Wallis and Futuna)+967 (Yemen)+260 (Zambia)+263
(Zimbabwe)Phone numberSendBy providing your phone number, you agree to receive a
one-time automated text message with a link to get app. Standard messaging rates
may apply.