On September 2017 I started studying a Master of Science in Artificial Intelligence in the UPC School of Barcelona as I was an outlier student I would like share my experience and maybe help someone to take decisions. What do I mean by an outlier? I had 41 years with 15 years of working experience most of the students where on their twenties and searching for their Phds. I had to quit from working, I had to quit from regular incomes, I had to quit from what I was doing during that 15 years and go to the Uni 🙂 yes, it sounded funny!.
If you are planning to study a Master in AI I know what are your fears but you will not read about them unless you go to the last points. OK it is difficult, really hard, I needed full time and dedication. During my studies I had been able to do it without having fears of a total fail I have never been interested in research and university has been 70% of my time studying and 30% having fun, sports or hobbies. Nevertheless, what I found in the Master has been a group of students 100% committed and passionate by their studies and their career as scientists, there were no parties, no sports, no holidays, no tourism (there were 14 different nationalities in my courses so lots of them where from different countries) just mathematics, science, computing, coding and work work work. My admiration and respect for this bright and young group of students without their 30% of time for fun.
I will not enter to review the UPC’s Master program or teachers, go and try it. I will try to give some tricks that worked for me for for succeeding. Those science masters involves lots and lots of mathematics (I mentioned your fears), science theory, engineering concepts and some coding in different languages mostly matlab, python and some Java. But, as a Master it also involves lots of team work and it is there where I focused my tricks for succeeding. Try to choose carefully your team and don’t be shy the first weeks where we find teams for each course, maybe you are not so good at coding or problem understanding but you can do excellent reports, maybe you can help with technological infrastructure or work coordination the team will be delighted with your job as they are future scientists or engineers and this is where they lack more of experience.
Be very attentive to how the teachers will grade you, they normally balance more for practical works than for exams but they all have their own system sometimes they need an algorithm for computing your final mark as they use lots of combinations, delivered work, final practical work, minimum marks in exams in example. Be strategist find a plan for succeed and be confident in following it (specially during exams time) you will find interesting subjects studying a particular course but you don’t have time to read about it keep it for another moment and focus in the course. I used for myself Project Management tools that I used in my career I had my own Jira and Confluence and put my great efforts in the key points for succeeding the course. Be confident in your plan even when you are days without good sleeping. If you are like me and have been lots of years working you will probably have much more experience in documentation, meetings, timings, milestones and keeping the deliveries accurate to the limits imposed by the teacher and you are probably good managing timing and stress you can do this tasks in the team work but you will learn nothing my strategy was to try to teach this skills to other team members and try to learn maths and science from them. It worked.
Final point about your fears 🙂 a science master of AI is hard in mathematics as everything is maths. Try to recover your basic knowledge of before finishing your studies, recall yourself in matrices,gradients, geometric functions, algebra, probabilities and a long etc… (seach in Quora there are very good advises in what exactly will it be needed) I hired a math student Julia for having private classes it was a good idea, don’t try to understand how a AI software works without having clear math concepts (just what you learned in college) and try to have access to those key concepts I founded very useful something like this cards.
As I read in François Chollet twitter
“Neural networks” are a sad misnomer. They’re neither neural nor even networks. They’re chains of differentiable, parameterized geometric functions, trained with gradient descent (with gradients obtained via the chain rule). A small set of highschool-level ideas put together”
So it’s reasonable to be afraid you will need those maths but it doesn’t have to stop to you to learn something so amazingly thrilling like Artificial Intelligence because if you are focused and constant you will do it and yes, you will have to code.
