What are Some “Advanced” AI and Machine Learning Online Courses?
Where can you find not-so-common, but high-quality online courses (Free) for ‘advanced’ machine learning and artificial intelligence?
Why this article?
Many young professionals, who have started their journey into data science, and machine learning, face a common problem — they have completed one or two basic online course, done some programming lessons, put up a couple of projects on Github, and then… then what?
What to learn? Where to find focused resources?
In one of my previous articles on Medium (published by the TDS Team), I discussed, at length, where you can find MOOC (Massive Open Online Course) for jump-starting your journey into data science and machine learning. That article assumed the reader to be a beginner and covers essential MOOCs, which are optimized for basic and intermediate learning. You can check that one here,
How to choose effective MOOCs for machine learning and data science?
Advice for professionals in non-CS field eager to learn and contribute to data science/machine learning. Curated from…towardsdatascience.com
I wrote another detailed article specifically focused on the topic of mathematics concepts you need to master for data science and machine learning and which courses to study. You can check that one here,
Recently, I have been receiving a lot of messages in my personal email and LinkedIn inbox, mostly from bright, young professionals, asking similar questions and my suggestions about online courses.
I mostly have a ready answer for those messages. I just send them a list of my articles (which, in turn, contains links and references to other highly cited articles from KDnuggets or Team AV). In most cases, I receive happy replies :-)
However, since writing those articles, I have personally taken few more ‘advanced’ courses in AI and machine learning (ML), seen discussions and reviews around few more, and naturally felt a need to update those references.
After some thought, I decided it is best to leave the original articles, as they really cater to beginners and have served their purpose well for many readers, and try compiling a fresh list of online courses.
That’s what this article is about.
What do I mean by ‘advanced’ level course?
Advanced’ is a relative term. It is best to have a baseline to explain the word in this context. Fortunately, we almost have a gold standard when it comes to ML online MOOC — Prof. Andrew Ng’s Coursera course (the original one, not the Deeplearning.ai specialization).
Therefore, by ‘advanced’, in this article, I allude to two characteristics, which need to be present (not necessarily simultaneously) in the courses that will be discussed,
- Significantly more breadth than the aforementioned course i.e. covering more advanced and diverse topics
- Highly specialized focus related to AI or ML
I hope I make it clear that my intention is not to say that Prof. Ng’s course is a rudimentary one. It is still the best introduction to the world of machine learning one can ask for — particularly for beginners. But, after you finish that course, do some programming, feel comfortable about the mathematics concepts, you should build on your base and learn diverse topics.
I just hope that this article can help you do that by listing some free MOOCs with that singular focus.
What is the singular focus for selecting the courses?
AI and ML are hot topics and there is no dearth of free online courses covering those subjects. Although, I have found that there are a surprisingly little amount of true high-quality AI courses out there.
Yes, I am in that camp, which firmly believes that deep learning is not artificial intelligence, and therefore reject the notion of any course, having the word “AI” in its title but covering nothing more than deep learning frameworks in Python, to be classified as an AI course.
So, to restrict my listing to a limited number of high-quality courses, I laid out some simple ground rule or filters.
- I tend to avoid any course with a significant focus on a particular programming framework/tool i.e. no course with a name like “Machine learning with Python…” (some examples or code snippets are fine)
- Following the same logic, the list will have courses with a strong emphasis on the theoretical foundation — this mainly favors university courses over those offered by individual entrepreneurs or companies (e.g. fast.ai, Google, Microsoft, IBM, etc.)
- Similarly, I included Udacity courses which are taught by university faculty or renowned researchers like Sebastian Thrun or Peter Norvig. I did not include their nanodegree references, which I do not find intellectually uplifting.
- I put included two topics which have enormous importance for true AI learning but receive less than usual attention — reinforcement learning and game theory.
- No course focused primarily on data science/data engineering/digital analytics/applied statistics. They all are critically important topics to learn in today’s world but I prefer to separate them cleanly from my focus on pure machine learning and AI for the sake of this particular article.
I believe this focus will automatically curate the list toward high-quality, foundational courses in AI and ML, which can benefit intermediate to advanced learners.
You will be the judge, after all.
Personally, I have not taken all of these courses although I finished a significant portion of them. So, I tried to keep my comments about the courses brief and factual.
The links and references
Without further delay, here is the list.
General machine learning and deep learning
These are courses covering general ML and DL topics.
- Georgia Tech’s “Machine Learning” course on Udacity: This is one of the most comprehensive ML courses out there with coverage of supervised, unsupervised learning, randomized optimization techniques (e.g. genetic algorithm), reinforcement learning, and even introductory game theory concepts.
- The original Stanford classroom version of Andrew Ng’s lectures: This is the full classroom version of Prof. Ng’s ML course at Stanford. Covers foundational topics of ML in depth which are missing from the watered-down online MOOC.
- “Advanced Machine Learning Specialization” by National Research University Higher School of Economics on Coursera: This is a great set of courses (5 in total) offered by Russian researchers. Good coverage of practical deep learning techniques along with foundational concepts.
- “Machine Learning at Scale” by Yandex on Coursera: Covers deployment and scaling up of ML models using MLib/Spark etc.
- “Machine Learning Caltech course”: This was on edX before but since moved to Prof. Mostafa’s home page. The link point there. It is a great foundational course in deep mathematical aspects of machine learning and learning theory in general.
- “Machine Learning Fundamentals” by UC San Diego on edX: A well-balanced course teaching core theoretical and practical concepts in ML with emphasis on algorithmic issues.
Artificial Intelligence and Game Theory
These are AI and game-theory related courses.
- Udacity’s “Introduction to Artificial Intelligence” course: The most comprehensive core AI course you will currently find on the web. It is taught by two renowned experts — Sebastian Thrun and Peter Norvig. They cover topics like — AI search algorithms, planning, representational logic, probabilistic inference, machine learning, Markov processes, hidden Markov models (HMM) and filters, computer vision, robotics, and natural language processing.
- Columbia University’s “Artificial Intelligence (AI)” course on edX: This is also a comprehensive review of essential topics in AI, but at a less rigorous level. This is a good introduction to the broad field of AI covering topics like — types and definition of intelligent agents, history of Artificial Intelligence, search, games, logic, constraint satisfaction problems, examples of AI applications in natural language processing (NLP), robotics, and computer vision.
- Stanford’s “Game Theory” on Coursera: This is a great introduction (yet comprehensive) to the wonderful world of game theory covering all essential topics like — Nash equilibrium, Mixed strategy, Correlated Equilibrium, Subgame perfect, Extensive form, Repeated game and folk theorem, Bayesian game, Coalitional game.
- “Knowledge-Based AI: Cognitive Systems” by Georgia Tech on Udacity: A comprehensive course on the traditional AI (or as they call it GOFAI) covering topics like — semantic networks, means and end analysis, case-based reasoning, incremental concept learning, logic and planning, analogical reasoning, constraint propagation, and meta-reasoning.
These are reinforcement learning related courses.
- Georgia Tech’s “Reinforcement Learning” course on Udacity: This is probably the most comprehensive RL course out there. Both instructors are extremely knowledgeable and passionate about the subject. The mode of delivery is conversational styled and fun. It covers the entire gamut of topics like — MDP basics, Temporal difference (TD) learning, value and policy iterations, Q-learning, convergence properties, reward shaping, Bandit problem, Rmax analysis, general stochastic MDPs, state generalization, POMDP, options, goal abstraction techniques, mechanism design, Monte Carlo tree search, DEC-POMDP, policy critic concepts, etc.
- “Practical Reinforcement Learning”, offered by National Research University Higher School of Economics on Coursera: This is another fantastically comprehensive course covering essential RL topics. The main difference from the Georgia Tech course is that it does not include game theory discussion and offers more discussion on Deep Q-learning instead. It is a more hands-on type of course teaching you practical tricks (but not necessarily full blown code) for building RL agents.
Other relevant topics
- Udacity’s “Artificial Intelligence for Robotics”: This is a great little course specifically focused on application of AI to the field of robotics, taught by none other than Sebastian Thrun. He covers topics like — localization, Kalman filters, particle filters, advanced AI search techniques, PID control, SLAM (simultaneous localization and mapping), etc.
- “Mathematics for Machine Learning Specialization” by Imperial College, London on Coursera: A great specialization of four courses focusing exclusively on building the mathematical base for machine learning. It covers — multivariable calculus, linear algebra, and principal component analysis (a full short course on that).
I hope to have given you some pointers on free online courses which cover somewhat advanced topics of machine learning and artificial intelligence. In this article, I specifically listed MOOCs and did not consider free-form video lectures (exception being the Stanford CS229 course). You can, of course, search for such video lectures from various university’s online platform and often they are of very high quality.
Wishing you great success in your journey to learn these exciting topics!
If you have any questions or ideas to share, please contact the author at tirthajyoti[AT]gmail.com. Also, you can check author’s GitHub repositoriesfor other fun code snippets in Python, R, or MATLAB and machine learning resources. If you are, like me, passionate about machine learning/data science, please feel free to add me on LinkedIn or follow me on Twitter.
Tirthajyoti Sarkar - Sr. Principal Engineer - Semiconductor, AI, Machine Learning - ON…
View Tirthajyoti Sarkar's profile on LinkedIn, the world's largest professional community. Tirthajyoti has 8 jobs…www.linkedin.com
Bio: Tirthajyoti Sarkar is a semiconductor technologist, machine learning/data science zealot, Ph.D. in EE, blogger and writer.
Original. Reposted with permission.
- Essential Math for Data Science: ‘Why’ and ‘How’
- Synthetic Data Generation: A must-have skill for new data scientists
- How Much Mathematics Does an IT Engineer Need to Learn to Get Into Data Science?