I am an international student. I wanted to specialize in Deep Learning, so I joined MEng in AI.
I can’t speak of undergrad, thesis, Ph.D. or other professors at UC. My duration here was roughly around 8 months, so take this review for what it’s worth.
I was at UC for 2 semesters. I had to choose 4 courses each semester, of which 1 would be non-technical like entrepreneurship, leadership etc. The 6 other courses are related to AI, no unrelated mandatory course which is great.
This is THE BEST course at UC for me. The course starts off with philosophy-like lectures about intelligence. The majority of this course is teaching the basics of neural networks. The professor explains the inner workings of neural networks in meticulous detail. The course also includes unconventional neural networks like spiking neurons and Hopfield networks, ensemble networks, etc. The exams are also good — few one-liners, some longer answers, sometimes open-ended questions that do not have a correct answer.
The professor is Ali Minai, the director of the engineering department. Ali Minai is one of the best professors at UC, a huge influencer for me. He is confident and knows his material really well. His words are very clear, he maintains that clarity of speech throughout the course. His style of teaching is in a manner that makes difficult concepts easy to understand.
You don’t have to take my word for it, read others’ reviews here.
This is the second-best course at UC for me, taught again by Ali Minai. Concepts here include evolution, emergent phenomena, etc. This class has so many open-ended threads that can be dug into their own rabbit holes. You have a lot of optional material to read which is really interesting. This course can open up one’s curiosity towards nature. I absolutely loved it. There are no exams, just one final project that you must present. Student groups come up with very fascinating projects.
One must not go into this course thinking you’ll get a good intro to deep learning. This is anything but. The course starts off with autoencoders and in each class, the professor takes the students through a new concept in deep learning. There’s a short but difficult quiz after each class. One must study the concepts well before attending the class or you’ll have a hard time following.
It’s taught by Yizong Cheng. I love the course, but I’m not particularly a fan of his. He offers nothing interesting to the subject. His materials are okay. He doesn’t explain much, just introduces us to the concept and talks in a hand-wavy fashion. The quizzes and exams are actually good.
Yes, UC has a dedicated computer vision course, but it is masqueraded with a different name: Deep Learning for Image Processing Applications. This is a Ph.D. level course that I chose outside of the curriculum because I love Computer Vision. The course starts off with the professor explaining how CNNs work. After a couple of weeks, students take over — each student voluntarily gives a presentation on any interesting computer vision papers. This is such a breath of fresh air for me, every class is exciting and new with a new teacher that is one of the classmates. There are no exams, just a final presentation.
The professor is William Wee. He is a very friendly teacher. He has a background in traditional image processing. He doesn’t explain the concepts with clarity, but that’s okay because students are the teachers here anyway.
The Textbook was the regular AI A Modern Approach which is good. Reading the textbook was fun and 100x better than listening to the lectures. However, you need to pay attention to the classes to pass the exams. The exam pattern is discussed before the exam so that’s useful. The worst part however is the fact that the professor uses an ancient logic language called FRIL. It is very similar to prolog, so there are some resources available online.
The professor is Anca Ralescu. She has a lot of background and experience in the field of AI. She boasts her connections with pioneers of AI research. Having a good deal of background doesn’t mean she can teach well. In my opinion, she is a terrible teacher (find professor reviews in ratemyprofessor for more opinions). She thinks in her mind that all of the AI concepts are so obvious and doesn’t consider the case of students feeling overwhelming. This happens in every class. The professor sometimes will not discuss the questions that appear in the exam, dismissing the questions as too obvious to answer.
Machine Learning class teaches 4–5 ML concepts like Linear Regression, SVM, and finally Neural Networks. This is a huge class, the classroom is like an auditorium and it’s usually full. This is again taught by Anca Ralescu, a huge bummer. I passed the exams by learning as much as I can online. Many students are too bored and fall asleep during the class and honestly, cannot blame the students because she’s just not an engaging teacher.
StartupUC / Special Topics in Entrepreneurship
This is one of the non-technical courses I took. I thoroughly enjoyed this class. Every student presents their own startup idea. Students are formed into groups to develop their startup further. The coursework is based on a book named Lean Startup by Eric Ries.
The professor is a former startup founder named Chris Petersen.
I attended a few startup events in Cincinnati. Turns out that Cincinnati has a good startup culture! I am currently working at a Cincy startup!
Management of Innovation
Pure waste of time, took it because it is just for 2 months.
I am an overly studious guy, occasionally attending events, so I cannot speak much of athletics, parties, and extracurricular activities. I attended a few programs, HyperloopUC, Salsa class, and a few networking and fun events which were all great. The administrators were nice to talk to.
Coming from a place where job opportunities are basically gifted to students — undergrad at KMIT, UC felt quite different. Companies don’t randomly set a date and show up to interview students, but they visit the campus on a single day — the event of Career Fair. Other than that, students occasionally get job alerts from the job advisor where one has to apply online if interested. So students are pretty much on their own when hunting jobs.
UC conducts a career fair every semester. About 90 companies come to the event, of which only 20 or so are software companies, almost none are AI-related. Most of them are construction, manufacturing companies, etc. It is nice to interact with the companies, one must plan to navigate well, taking full advantage of the guide book given. I got my first 2 jobs in the first semester’s career fair. Not everyone is as lucky as I am though, I know countable friends who got the job from fairs.
The job advisor Amanda McLaughlin is a very kind person, helped me build my resume and more.
When the COVID pandemic hit, I was halfway done with my second semester. The biggest bummer for me was that I couldn’t meet the professors and students in person. If I had to watch online lectures, I could’ve done it back in India. The relationship with professors and in-class experience is why I go to the university. Now that’s lost.
Classes however recovered quickly. The Materials for all the courses were already online since day 1, so only lecture videos and live lectures had to be introduced. I wouldn’t say I disliked the online lectures, but there was a definite disconnect with the professors and students.
UC provides off-campus housing for graduate students. The apartment is very neat and overall very good.
The university is approximately 3 miles away from downtown and 1 mile from the Cincinnati Zoo. The surrounding area is urban and neat with mostly safe areas.
University has provided $10k as a scholarship for all international students.
MEng in AI at UC has been an amazing experience for me. The best part is the reduced course duration of just 2 semesters enabling focused coursework and opening the opportunity to work right after the coursework.
If I missed out on any aspects, please let me know in the comments!
So why was I interested in UC?
- Masters in AI. There were only 4 universities in the US that offered Masters in Artificial Intelligence. One of them is UC. Other 3 — Northwestern, UGA, and Northeastern University.
- Relatively low fees — $20k
- 1-year non-thesis course
A LOT of students came to UC from my city, nobody I know was rejected. So what is the real acceptance rate?
How to select a university for your master's?
Here’s what I did — only applies to CS students.
- Go to http://csrankings.org/
- On the left, select the areas that you are interested in.
- On the right, you’ll see a list of universities sorted descendingly by the number of publications and faculty.
- Assuming that a number of publications correlate to a better university, this will give you an academically personalized list of universities.
- However, you still do not know which universities are likely to get admission from. For this, there is no real statistic. I relied on the mean GRE scores provided in https://yocket.in/
- Once you’ve shortlisted about 50 or so universities, now the next step would be to go to each university's website, look at the course descriptions, fees, etc.
- The next step is to look at who is teaching these courses and find their reviews on https://www.ratemyprofessors.com/. You will now be able to shortlist to a small enough number to finalize the list!
But why am I interested in AI in the first place? That’s for another blog!
Thanks for reading! Checkout my ML website at https://ml.gallery/