GEM VEnturer: A Hidden Study Abroad Alternative

Also a Review of Tsinghua University’s Big Data and Machine Intelligence

Last summer, when my friends were busy lamenting over their exchange module mapping woes, I was furiously making plans A, B, C… Z for my upcoming STARS season. As an upcoming Year 3 ECDS student at the time, I was eager to finally register for SC4000: Machine Learning to gain more knowledge and experience in ML models. After my Year 2 internship, I had realised how highly sought-after machine learning skills were… anddd also how much I lacked them, so I wanted to bridge that gap as soon as I returned to NTU in Semester 1.

Unfortunately for me, SC4000 is only offered in Semester 2. Just when I thought that was the end of my plan, an email landed in my inbox. An email I had, to be honest, always ignored.

For the first time, I actually read it. Turns out, I could actually take some virtual modules at NTU’s partner universities overseas and clear some credits through the GEM VEnturer programme. After a bit of browsing, I found a virtual module at Tsinghua University titled “Big Data and Machine Intelligence”. This was it – the perfect opportunity for me to clear SC4000 without having to wait till Semester 2!

But… what’s the catch?

Hold on! Before we dive straight into the pros and cons, what even is GEM VEnturer?

If you came across this article title and thought to yourself: “Huh, GEM VEnturer, now that’s something I’ve never even seen before!” – then rest assured, you are not alone. In fact, when I told my friends that I was taking a GEM VEnturer module last semester, most of them also had no idea what it was 🥲 However, despite being relatively overshadowed by NTU’s other exchange options, I think GEM VEnturer is a real hidden gem for those of us who may not have the chance to go on the usual exchange programmes!

In short, GEM VEnturer is a programme that lets you take virtual or blended modules with students from other countries. There are mainly 2 options:

Global MOOC: You will take 1 virtual module (via Zoom or similar platforms) from NTU’s partner universities overseas while still studying in NTU as usual. Think of it as a partial SUSEP, except it’s only 1 module, and the module is from a foreign university.
NTU COIL: These are special hybrid NTU modules conducted by NTU and overseas professors. In these modules, international students will join you virtually, and you may also have the chance to go on a study trip overseas!

Moreover, as with other exchange programmes, GEM VEnturer modules can be cleared as MPE or BDE as a Pass/Fail course.

Since I’ve personally taken Global MOOC modules but haven’t tried NTU COIL yet, this post will mainly cover my experience with Global MOOC. If NTU COIL modules are more up your alley, do check them out here!

My GEM VEnturer Journey: A Personal Experience with Big Data and Machine Intelligence

Now that we’ve covered the basics of what GEM VEnturer is all about, let me share my own experience with the programme. Last semester, I took my first Global MOOC module, Big Data and Machine Intelligence (BDMI) from Tsinghua University, and this semester, I’m continuing the journey with Game Theory from the same university. Here’s a more detailed look at my experience so far, from signing up for modules to what I’ve learned along the way.

The Sign-Up Process:

Signing up for GEM VEnturer was a bit of a process, but nothing too overwhelming fortunately. First, I skimmed through all the modules offered by NTU’s partner universities for Global MOOC, and let me tell you, there were A LOT of modules.

Note: These are modules offered in AY24/25 Sem 2.

Not only that, the modules also span a wide range of topics from maths, sciences, computer science to economics, finance and even languages!

At this point, I was eager to apply for Tsinghua University’s BDMI course to map it back to SC4000: Machine Learning. Comparing the syllabuses, I could clearly see the similarities between the course outline and topics. However, I had to check and make sure this course mapping was possible before I committed to the course, and what better place to do that than in the Coursefinder portal.

*GEM VEnturer course mapping records are under the GemX – GEM Explorer options.*

*GEM VEnturer records are denoted with Vse to differentiate from GEM Explorer records.*

*And there they are, approved records for mapping BDMI to SC4000.*

Now that I knew it was possible to map BDMI to SC4000, it was time to submit my applications to BOTH the host university, Tsinghua University in this case, and to GEM Mobility Portal. Each host institution has a different portal, but the application process is generally seamless.

After that was the waiting game. The deadline to apply for Global MOOC modules was just before the Add Drop period, so I chose to register all my modules on STARS first in case I couldn’t get BDMI. The result came out on the Wednesday of Week 2, a few days before the end of Add Drop, so I had the chance to weigh my options and decide whether to accept or reject the Global MOOC offer. Of course, I decided to accept and try it out, and even managed to switch my indexes around to slot BDMI into my schedule more easily.

Planning For Assignments:

Now that I had successfully registered for BDMI, it was time to actually face the music and attend the course.

Luckily for me, I didn’t have to deal with any time difference as Singapore and China have the same time zone. This made attending the weekly lectures and keeping track of my assignment deadlines super easy!

As my host university was Tsinghua, my course timeline was a little bit different from NTU’s semester timeline. Most Chinese universities start their semester almost halfway through our semester, so most of my major assignments for BDMI came way after my major deadlines for NTU modules.

Source: BDMI Assignments
*My assignment deadlines were NTU’s Weeks 10, 14, and 2 weeks after Final Examinations.*

This made planning my studies and homework much easier, as I essentially had one less module to worry about for half the semester. However, they also end the semester later, so while my friends were busy hanging out at Gardens by the Bay on Christmas Eve, I was busy presenting my final project for BDMI.

…Well, we win some, we lose some, am I right?

Keep Up to Date with the Latest Technologies:

Learning from professors at another university was definitely an unforgettable experience for me. Most interestingly, while we had a main instructor for the course, he was not the only instructor we had. In fact, there were many professors who all collaborated to teach us different topics and shared about different machine learning techniques based on their own expertise. They made an effort to bring in the latest tech breakthroughs and demonstrated how to leverage them for our projects, which made the learning experience that much more exciting. It’s not every day I get to dive into the freshest topics in such a fast-evolving field!

Learning Through Projects:

As someone who always prefers to apply what I’ve learnt in projects rather than take exam papers, one of the reasons I wanted to take BDMI was because the mode of assessment was entirely through hands-on projects.

Throughout the semester, we were given some pretty interesting datasets, like historical prices of stocks and spectrograms of audio files, which helped us learn how to deal with different types of data efficiently. We were given a lot of flexibility to try out different machine learning algorithms and implement our own strategies to improve the accuracy of our models. While the process was definitely challenging, it was sooo satisfying to go from stumbling my way around in the first few weeks, trying to understand new techniques to realising how those techniques could all come together to improve my model’s accuracy later on.

*My first assignment: Predict stock price movements for 4 different stocks using various machine learning algorithms. It was interesting to see how each model had its pros and cons.*

*Throwback to my spectrogram classification model! Yes, my final results were rather abysmal, but considering my first model had an accuracy of* **4% 💀***, this was a win in my book.*

While our first 2 assignments were relatively structured with assigned datasets, the final assignment was completely open-ended. We could, essentially, do anything. Some of us identified a problem and suggested how to use machine learning to deal with it, some wrote a critique on a recent paper related to machine intelligence, and some participated in a competition to build a quantitative model to predict stock movements. At the end of the course, we all had to present our projects, and I found so many unique ideas from my classmates to learn from.

*For my pair, we decided to work on a Deep Learning model to predict mathematical misconceptions. It was tough but also super rewarding to see our final project come to fruition!*

Throughout the course, there was a lot of emphasis on learning through doing, with class discussions being a key part of the module. While there were plenty of optional and compulsory projects to try my hands on, the workload was definitely manageable, and I never felt too overwhelmed, which was a pleasant surprise!

Looking back, did I like my experience?

The simple answer is… yes! There’s a reason why I decided to take another Global MOOC module again this semester after all 😉 The course gave me the chance to gain new perspectives and learn from professors with different teaching styles, all while staying in Singapore. It wasn’t a huge time commitment, so I was able to focus more on my other modules. Plus, I managed to clear an MPE with a lot less stress about my grades!

However, just because my experience was mostly enjoyable doesn’t mean the programme didn’t have its flaws. In fact, there are a number of pros and cons you should consider before taking on a Global MOOC module:

Pros	Cons
More cost-effective than other exchange programmes (especially since NTU waived the fees this year)	Limited options, mostly from East Asia universities
Many modules can be mapped to BDEs and even MPEs	Limited number of modules available
Easier to plan – only need to slot in 1 module instead of managing multiple modules while overseas/on SUSEP	Can only take 1 module per semester and a maximum of 3 times throughout your degree
Opportunities to make new foreign friends	Hard to plan ahead as course listings change every semester
Option to audit courses for those who want to learn without the stress of grades	Possible language barrier

So what’s my advice?

Simple, just try it out! When the next round of applications for GEM VEnturer opens up, just take a look at the course listings to see what’s available and what works for you. If you can’t make up your mind, just apply! You can always turn down the offer if it doesn’t work out in the end.