2.5
Quality4.1
Difficulty37%
Would Retake98
Reviews37%
Would Retake
98
Reviews
Rating DistributionOfficial
5
31
4
2
3
5
2
2
1
58
What Students Say
“Lecture covers 5% of what's needed for homework”
CS472 - 1.0 rating“Took 7-week Big Ideas in AI course”
CSCIUA74 - 1.0 ratingClass Info
Online Classes
100%
Attendance Mandatory
49%
Textbook Required
0%
Grade Predictor
Your expected effort level
Predicted Grade
A+
Grade Distribution
Common Tags
Rating Trend
Stable
-0.13 avg changeRatings by Course
DSGA1008
4.4
(14)CSCIGA2572
3.9
(26)CSCIUA74
1.4
(5)CS472
1.4
(50)CSCI74
1.0
(3)Difficulty by Course
CSCIUA74
4.6
CS472
4.5
CSCI74
4.3
CSCIGA2572
3.7
DSGA1008
3.5
Reviews (98)
Deep Learning, Spring 2020. Lectures are interesting and easy to understand. You can find them on his YouTube channel. Coding assignment were too easy. Mid-sem exam were quite difficult. Project was too much work, too little time. Piazza discussions were fruitful. Engagement is encouraged. Grading was lenient. No office hours.
Terrible. Took 7-week Big Ideas in AI course. People constantly begged for our quiz results. He tells us to go ask someone else. The quizzes are insanely specific and hard. There is absolutely no guidance on your grade--no one knows anything. He and the TA's never respond to emails and when they do it's to ask us not to talk to them.
DO NOT TAKE. Lectures are on NYU Stream. Q&A sessions mandatory. Grading criteria vague. YET, they threatened to fail half the class week 2. Weekly quizzes are awful, in difficulty and in specificity. When begged for quiz feedback, they lie to us. Alf passes blame off. Hes created a digital hierarchy among students. A harrowing experience.
Very unorganized class. Class has gotten to the last week and students still have 0 feedback regarding their standings (quiz grades, attendance grades, disc post grades). Quizzes are unnecessarily technical and difficult for a P/F 2 credit class. When students ask questions regarding their grades they get morally attacked by Prof/TAs. DO NOT TAKE!
Don't take this Big Ideas course! The instructor provides 0 feedback on quizzes, despite promising to return grades all the time. He interrupts the guest lectures who are actually interesting, and doesn't allow students to speak in the Q&A sessions unless he needs clarification. Waste of 2 credits.
Disagree with 5 dif ratings because this course is about AI, a technical field, instead of the ethics of AI only. No equations involved. About 2hrs of video lectures every week plus Q&A 1hr and quiz about the lecture, so workload not light. About 1 or 2 questions each quiz are unclear. Great content for introduction to many fields of AI though.
Alfredo is one of the few professors with humor and hard-working spirit. During his labs, there are countless times that I have smirked. He certainly is creative on generating metaphors and making the dry and tedious material into fruitful ingredients, reminding me about mom's recipes. Please do not hand in your homework late. I learned my lesson.
I took the Deep learning course in Spring 2021. He puts up lots of effort in his labs and makes sure everyone understands it. The class is not easy to do and this time homeworks were really good and time consuming. You will get to learn a lot. Only take the class if you have prior experience/knowledge in ML or AI.
Very motivated and caring professor. Feels strongly about how teaching style (eg never shares lecture slides until after class), and teaches well. Tries to inspire interest and teach more than just what you'd need to pass a test (there is no exam) and provides many additional resources for you to learn more. Takes feedback well. Highly recommend.
Awful prof. In nyu I faced.
Not respect student.
Worst class I have taken in NYU. The homework and group project are unlinked with class material.
His "playfulness" wasn't welcome by overanxious students expecting a more bland professor (though this could be a response to clueless answers we give for participation pts) Assignments required external readings. Lectures do cover many interesting topics. Asks for feedback but prob irritated pedantic students who didn't get his sarcasm.
One final group project that was most challenging assignment, but no exams. Homework tough but they always grade leniently. You can get extra credits for almost everything if you put in some more work. Participating for the sake of participating was really annoying as we are graded on posting things online (there are better ways to do this!)
Online. Homeworks were ok but not always fully covered by lectures. Yann sometimes gives classes but those tend to be conceptual and about recent topics, not about theory Alf covers. Annoyingly, some students used their own work resources (GPUs, cloud servers, etc) to compete for the final project which greatly skewed results in their favor.
Required participation is demanding but it makes you focus. Also, TA, prof and fellow students are very active online to help. Lectures are not math-intense but are really intuitive to understand. HW is doable and further helps learning. Some demo in class are not posted but otherwise lectures are perfect. Prof is hilarious, helpful, and kind.
Okay, so some may say his course is interesting or his slides are nice whatever. He is not a good professor. Use too many unnecessary analogy in courses, give weird description to different terms and make things too complicated. he might think he is funny, but hes not Acknowledge he puts effort into teaching, but does not make students life easier
Agree with the guy below. His slides and descriptions are weird and not informative enough, neglecting the principle behind each terminology, which is not helpful for learning, and he is even proud of his slides.
Professor Canziani is pretty far from a good instructor. He was very slow to answer questions on CampusWire, his lectures were unclear, and he skimmed over the tough stuff, which left us struggling with the coding assignments.
Just a Joker
Let's just admit that he is beyond terrible, confusing lectures + hard lab + never answers campuswire questions in time, he literally ruins the AI course experience which was quite good when it was instructed by Ernest in the first half of the semester.
Strangest Professor in DS department. He knows things but not a good lecturer and assignment-maker. Ernie did a much better job in the first half. Alf made assignments unclear so students, with years of AI research, couldn't understand what he was up to and he answered questions with something unhelpful that we already knew and considered.
Not able to teach, but He is full of confidence in himself
So bad at teaching difficult concepts. Homework is awfully worded with unclear instructions & buggy code. He disappears right before our assignment due and now even the final exam. Campuswire is full of unanswered questions about final/homework but he's been missing the entire time.
Unclear lecture slides & assignments, not accessible, confusing lectures. Really enjoyed the first half of the course with Davis, but went downhill with Canziani. Still haven't received any grades from the second half of the course & the final is tomorrow.
Dove sei?
What can I say? Mamba out!
I don't even know how to start describing how awful he is. If you choose dancing and tweeting over answering our questions before the final exam, maybe teaching is just not the best job for you. The exam questions are worded so atrociously vague and confusing that I began to question whether we are even speaking the same language. Unbelievably bad.
Not that helpful
He needs to learn how to speak English first before giving us assignments or exam questions.
The only professor at NYU whose slides reminds me of 3Blue1Brown. He puts in tons of effort into preparing the material and often extend class time to cover unfinished stuff. The only con is that the deadline can be very tight relative to the hard material, but he really knows how to teach and engage from grad students to middle school kids lol.
Hey Alf, try something else than teaching!
A lot of people are being disrespectful, but I understand why, we are all frustrated by his responsiveness out of class, and I thought his lectures/slides weren't good from the beginning. I think he's friendly and easy to approach but will probably never reply on campuswire, and ends up giving impossible questions without teaching the material.
WORST NIGHTMARE. Unable to explain anything. Buggy programming assignments templates. Zero feedback. Never answer students' questions. Take him if you also want to have such a fun experience.
I don't know how to learn in this course. Bad organization for slides, errors in assignments, talk badly over slides and giving very vague guidance for learning the material. Avoid.
Bro thinks he gets more views on Zoom than Twitch. Go back to Twitch my guy :((
Bro should stick to anything but teaching lectures especially in AI. Even the AI will not be able to understand him. Jovial guy but hard understand ANYTHING he teaches.....
Aesthetic slides but not helpful slides. Programming assignments were confusing. Written assignments are tough. Test was awful. Still learned something but it was tough overall.
I kind of understand why he didn't pose assignment solutions before final. Turns out his final basically has nothing todo with his assignments. There is a list of questions that turns out to be exteremely important, but there are only questions, not answers nor examples.
impossible
Dear Alf, enjoy the joint complaint from us to the department :))
Wrote a confusing, poorly worded, unclear final exam and released none of the grades of assignments pertaining to the second half of the course (the majority of the final content) until after the final. Created a campuswire and had very few to no professor or TA answers to student questions. Not bad at explaining content in class though.
He's a nice guy, but a poor teacher. His explanations were winding and hard to follow, and he expects a level of understanding that is not even close to taught in lecture. It's like he's dumbing things down to the undergrad level but then expecting the graduate level of understanding despite not teaching it.
Waste tuition
choose him if you want to regret in the future
Active on twitter and everything else but on campuswire or emails where students are begging him to answer questions. If you care about being a social media influencer more than a teacher, just go get a different job outside academia. NYU simply needs to stop hiring people who consider the professor job as a shortcut to becoming a popular YouTuber.
Worst professor ever. Very condescending, refuses to answer even simple questions, cares more about social media fame than actually teaching. Treats class like a Youtube stream instead of actually teaching, refuses to help and instead blames students for asking the questions. Acts as if he does everything right even with students begging for help
Alf is a good professor. It lets students experience the cruelty and absurdity of reality and raises their awareness of fraud so that they are better prepared for life.
THE WORST!!
special shoutout for making us write 15 gradient descents on the final. What a deeply enriching question! I now feel as capable as an 8-core processor, overheated and questioning my life choices.
I still cannot process the fact that there even exists a professor on this planet who didn't reply to any of the students' questions about the final exam until the second day after the exam. He also did so with such a condescending tone and a grammar mistake saying "there are only a finite number of minutes in a day" with no sign of apologizing.
I firmly believe that Professor Canziani should be dismissed immediately. It seems that the department is unaware of his conduct in class, or they would have taken action already. His behavior is highly inappropriate; he is irresponsible, dishonest, condescending, and aggressive. Someone gotta let the department know what we have experienced.
He is the AI system that Ernest created to teach students but unfournately doesn't work.
Who answers questions about the final exam 24 hours after the exam is finished? Canziani does! With such a dismissive tone he even blamed students begging him to address our inquiries before the exam. Lecture covers 5% of what's needed for homework. Basically expecting you to build a rocket after he taught you Newton's Laws, and it was bad teaching
Tough grader and gives difficult assignments without explaining the material for the assignments.
Hey man, find another job.
Thank you for ruining my final week's schedule
Bungled the situation surrounding our final exam + lack of communication. However, he is indeed goated with the sauce and has offered leniency for our last coding project with bonus opportunities.
The previous comment has made me doubt myself for a second that I wasn't grateful enough, but sorry, I'm not gonna take take the last-minute bonus on extra work two days before the semester ends. Students should not have to "camp outside" their professor's office to get all the necessary information they need to do well, if that's even possible.
Ah, now it all makes sense why Alf gets to do whatever he pleases before the final. Because there's always a line of people ready to shower him with gratitude the moment he tosses them a tiny scrap of kindness, right?
Disgusting sauce!!!
Bro, he is so confusing. You have to take him with the another professor but just decide ur pros and cons cause he is not it. Good Luck! But like how did the whole class fail the final (which he taught) and everyone passed the midterms with flying colors (which Davis taught), its just not making any sense.
Despite some initial misconceptions about his teaching and his material is quite challenging and abstract, it seems that this is the nature of what we're learning. For the final programming assignments, he gave students extra points for finishing all sections to their overall grades and was willing to help students complete the work together!
Not worth
Contrary to all the negative reviews, he is great. He really cares about students' understanding of the material. He gave us over a month to complete the programming assignment (plenty time), which many of these negative reviews are about. The final review he gave for the exam was very identical to the final, I'm not sure why people are mad.
How can anyone call a professor who ignores all forum questions and emails until after the final exam great? Preferring an irresponsible professor is a personal choice, but plz don't mislead students who genuinely want to learn. Someone claimed we had a month for the assignment, but 85% of the material was taught less than a week before it was due.
Worst prof possible. Never take under ANY circumstances
I am quite surprised to see all the hate towards Prof. Canziani on RMP, when in fact he's just a chill guy who puts in a lot of efforts towards his teaching (just look at this slides, notebooks, videos, etc. available online). I agree some may not be used to professors with a funky personality such as his but that was actually refreshing tbh.
The class is very challenging at the beginning, but it becomes clearer and more intuitive over time. The class materials are highly illustrative. Alfredo is very patient when answering questions and doesn't mind explaining concepts from the beginning. It could be challenging if you have no background in ML. Highly recommend
AWESOME PROF
AWESOME PROF !!! WORTH IT
An Interesting professor. I saw many bad reviews about the final exams? But for DSGA1008, we never meet that questions, I feel everything is good. Also if we can have more practical DL stuff in class, it will be better
I am taking DSGA-1008 this semester and I found professor Canziani very funny. He gave very interesting lectures and he loves to use some funny stories/quotes to inspire students. I would definitely recommend him to everyone who wishes to take Deep Learning courses.
Good professor. Very interesting perspective of Energy based models taught in the course.
Prof Canziani works hard on his lectures and slides, and genuinely cares about whether students understand the material. He's approachable, funny, and engages with the class in a way that makes learning more enjoyable. Unfairly rated, he's a great teacher who truly wants his students to succeed!
Enjoyed the course. I spent a lot of time to understand the concepts and it is rewarding.
Fortunately didn't have final exams since it's a grad course; so can't relate with a lot of negative reviews of Prof Canziani. But as a teacher, I feel he did a fantastic job and does not deserve such negative reviews.
Prof. Canziani is amazing, he takes feedback very seriously and explains material in easy-to-digest ways so that everyone understands. His teaching made this course fun and rewarding--he's awesome :)
I really enjoyed Alfredo's Course, he is very funny and very nice. He is always available for question answering if you go over to his office. He often have funny references in the lecture, he uses many metaphorical words to make Deep Learning easier.He cares about the student's understanding, keeps encouraging us to ask question in lecture!
I took Deep Learning 24 Fall. He is passionate about teaching and makes sure you understand the details of the course. The homework is quite doable and each took me 1 day to complete, and the grading is generous. For deep learning he is underrated. Personally I think he's funny and love his pink hair. Seems to get pinker throughout the semester.
Alf's pretty good at explaining things, but he really needs to get his course sorted out. His homeworks are pretty lengthy and the code needs an unintuitive jump (or maybe it doesn't if you've also worked at Meta or something) from what's taught in class. Also the final project was a bit of a mess. And hold some goddang office hours.
Alf might be a good prof for phD and experienced researchers but he is terrible prof for students with limited background in ML. He insults students and constantly act like he is smarter than everyone in the room including Yann Lecun. Most shockingly, I never seen a professor sending an email begging students to give him good RMP ratings lol
Alf stop buying water army that uses the same language in every single new comment with a constant high rating of 5, and stop downvoting every negative comment. Just so fake. If you don't understand why there's so much hate directed towards you, then read the comments carefully, retrospect and try to become a (much!) better instructor and person.
I don't write RMP ratings much but I think current RMP ratings of Prof. Canziani is so misleading. Strongly recommend taking his graduate-level Deep Learning course especially if you want to do research or pursue an advanced engineering work in the AI&ML field.
Trust me, he is the worst
Avoid taking a class with Canziani at all costs!!! Despite putting in a lot of effort he is a horrible teacher and lecturer. His "simplified" notation and visuals makes it impossible to follow. Almost impossible to link what he is saying to other reference material. I like LeCun's approach to DL, but Canziani will not help you understand it.
He puts way to much effort in making sure that students understand a particular concept. His terminologies and way of teaching is phenomenal. His books and texts are unmatched. Take this course if you actually love the field, you definitely gain something out of it.
Professor Canziani is an exceptional professor who actively encourages students to think critically during class. I appreciated his engaging teaching style andway he presents material. I hope he continues to teach with the same energy and passion, even if some students may not immediately recognize the value of his approach.
Professor Canziani is one of the best instructors I've had. His classes are insightful and conceptually clear. I would highly recommend taking a class with Professor Canziani if you're looking to strengthen your conceptual foundation and gain a deeper appreciation of the subject. It's an experience that stays with you well beyond the classroom.
Bro seriously needs to fix the broken coding template in these assignments. Did you just let some outdated LLM spit out the whole thing and never bother to test it? There are countless ways to specify a python environment yet Alfredo somehow managed to ignore every single one and dump a mess of untested code on the students to figure out.
Professor Canziani, is one of the best professor's I've had at NYU. He actually explains and writes things on the board. He's very smart and full of knowledge. Yes, the class is hard but that was to be expected for the quality of material being taught. He has made me look at ML models in a different way by visualizing them. 11/10
Fun fact, Alfredo has the lowest rating on RMP among all teaching professors in NYU's Math, Data Science, and Computer Science departments. After taking his course, I'd say it's well deserved.
This reviews are misleading, I believe the the Deep learning course is really great and is structured in a unique way wrt to energy based models and prof. is really knowledge able especially the pictorial representation of various topics. I would highly recommend prof. Alfredo
The class was a bit unstructred in terms of grading and has an online oral exam instead of written test. no midterm but regular surprise quizzes which is 30% of the grade and remembers all his students and reached out individually if u participate in class and ask questions on google chat.
Alf is an amazing professor that really cares about you understanding the process rather than just processing information. He really cares about reasoning and his course really taught me how to approach any DL framework. Please do not use Ai for his HWs as this will offend him. If you put in the work then you will learn a lot. I would recommend it.
Canziani is a really good prof because he understands how to lecture and teach well so you can really learn a lot just by going through the class. More importantly though, he will put so much effort in to meet you in the middle if you put effort into his class which is rare these days with professors. Would recommend to take his deep learning class
Amazing intro to DL. I'd sign up for pt2 immediately if offered. Superb clarity: high quality slides, engaging and clear instruction; pedagogical framework is unique and well suited (EBMs). Great selection of material in a crowded, low SNR field. Multiple landmarks explained with both depth and intuition. Insights and intuitions transmitted CLEARLY
This is a really good course, though it may not be for everyone. If you're interested in understanding the theory behind practical methods - how ideas are motivated and developed - this course is definitely worth taking. Alf and Yann put a lot of effort into prepping the material and genuinely care about making it a good learning experience.
Class Info
Online Classes
100%
Attendance Mandatory
49%
Textbook Required
0%
Grade Predictor
Your expected effort level
Predicted Grade
A+
Grade Distribution
Common Tags
Rating Trend
Stable
-0.13 avg changeRatings by Course
DSGA1008
4.4
(14)CSCIGA2572
3.9
(26)CSCIUA74
1.4
(5)CS472
1.4
(50)CSCI74
1.0
(3)Difficulty by Course
CSCIUA74
4.6
CS472
4.5
CSCI74
4.3
CSCIGA2572
3.7
DSGA1008
3.5