**Statistician Showcase**

**Jump to Article:**

**Young Statistician Profile: Ulrich Kotze**

**Young Statistician Profile: Ulrich Kotze**

20 August 2021

Ulrich Kotze is currently busy with his Master's degree in Statistics, and has competed in a number of hackathons over the past few years. SASA spoke to him about his interest in statistics, and what he has gained from his hackathon experiences.

**Tell us a little bit about yourself.**

I am in the final year of my Master's program, and I am also a junior lecturer. Besides my passion for numbers, I am an avid trail runner and mountain biker. I love to spend my free time in the mountains and push my limits. Statistics is a demanding degree to pursue at any level. Always having a race to train for or an adventure to look forward to helped me to escape the pressure somewhat and keep my mind fresh.

**How did you become interested in Statistics?**

I am pretty lucky that I have an older brother who happened to pursue a degree in Statistics. I initially made a point that I wasn't going to study the same thing as him. But, the more I listened to the stories of the cool stuff they got to do at work, the more my interest got sparked. Being in high school at the time, my brain couldn't fathom how "machine learning" was capable of all of these incredible things. It sounded like some form of magic.

From a young age, I was always interested in Mathematics and programming. I decided to take IT as a subject in high school, which was my introduction to programming. Although it was in Java, I still believe that those three years of programming always gave me an edge in the Statistics classroom. I have always loved mathematics. I just didn't necessarily enjoy Math classes. I don't think I was the most straightforward student to have in class. I was forever staring out the window but somehow always managed to have the correct answer when the teacher caught me not paying attention. I always knew Mathematics was necessary, but at the time, I just didn't know why. My daydreams often involved figuring out how the classwork applies to the real world. I didn't realise how important those years would actually be in getting me where I am now.

To answer the question, there was never really a specific moment or thing that led to my interest in Statistics. It was instead a build-up of many small things that eventually led me to where I am now. Statistics is the branch of Mathematics that solves real-world problems. We are not mathematicians, but we are also not scared to get down and dirty when we need to. As a statistician, I believe we like to ask the question: "How can this help me solve a problem?". So, I guess my pursuit of Statistics finally helped me clarify those things I dreamed about in Math class. It helped me understand how the "magic" works and how I can use it to solve real-world problems.

**Where did you study and why?**

I attended Paarl Boys High School from 2011-2015, after which I enrolled in a BCom degree majoring in Computer Science at Stellenbosch University. During my first year, I realised that Computer Science at University is not necessarily what the entertainment business sells. It wasn't all building the next Facebook and learning how to hack companies as the naïve first-year me thought it would be. Luckily the structure of the BCom degree at Stellenbosch University allowed me to quickly pivot my majors to Statistics, Economics and Quantitative Management. In 2019 I pursued an Honours degree in Statistics also at Stellenbosch University.

**How did you become interested in taking part in a hackathon?**

As a student, we need some way to build a bit of credibility – to show that all of the knowledge and skills we have acquired is useful and that we know how to use the skills to solve problems. We thought that a hackathon would be ideal for getting some exposure to people in the industry and showcasing our talent. These days it seems that a degree is not enough and that you need some way to separate yourself from the rest of the pack. Showcasing the quality of your work in a hackathon is one way to do this.

Talking to the winners of the year before, they highly recommended it and said it would be an enriching experience. I distinctly remember them saying that you are guaranteed to learn more than you would in any module, whether you win or lose. This couldn't be more true. Hackathons almost always come with a problem statement that does not have a conventional solution. It requires creativity, and as a university student, it often requires skills that you have not been taught yet. Taking part in hackathons and doing well showcases your ability to interpret problems and develop unique workable solutions. The solution you come up with is also your own work, so you can show this to prospective employers as an example of the type of solutions you can come up with.

**Could you tell us a bit more about what goes into taking part in a hackathon?**

I have taken part in three hackathons, winning two of them. All three had different formats and time spans. There are two standard formats. The first involves a problem statement and a dataset. You must use the dataset and usually some form of additional data source to solve the problem. At the end of the hackathon, all the teams need to present their solutions to a panel of judges. The judges are usually a combination of the sponsor's representatives. The presentation is generally relatively short, which is quite challenging when you need to summarise a few weeks of work into 15 minutes. After hearing all of the presentations, the judges deliberate and choose the winner. The second format is similar, but the judging requirement is more straightforward. The best model wins. You receive a problem statement and a training dataset. You are asked to make predictions on a test dataset, where the ground truth is held by the organisers of the competition. You may use the training data in any form you see fit to train your model. You then use the model to make predictions for the test dataset. The predictions are submitted online, where they get evaluated. An evaluation metric such as Area Under Curve is usually chosen to rank the teams. The team with the best score wins.

Hackathons are usually team competitions and take place over a weekend or a few weeks. Personally, I am more a fan of hackathons with a presentation at the end and require a degree of finesse and strategy. Hackathons, where one evaluation metric is the sole judge, is a bit different since it feels like you are strangling the data just to improve your model by 0.001% and in the process violating assumptions that would make Ronald Fisher turn in his grave.

Hackathons that are being judged by a panel require a degree of finesse and know-how. In both hackathons that we won, we spent a significant amount of time deliberating the question and how we could create a solution that would please the judges. The question is usually based on the context of the title sponsor. You are not required to build a solution for the sponsor, but you are guaranteed some brownie points if you do.

**What are the benefits of taking part in a hackathon?**

The benefits are twofold, firstly you are guaranteed to learn some new skills. These skills could be learning a new form of analysis or presenting to an audience, or seeing the creative solutions that the other contestants come up with. Secondly, if you do well or are happy with your work, it allows you to showcase your work to potential employers.

There is also the added benefit of some nice prize money, which comes in handy to struggling students. It is definitely a nice incentive to keep you going in the early hours of the morning.

**What advice do you have for fellow students wishing to take part in a hackathon?**

I would advise students to not get too obsessed with winning. I think the goal of taking part shouldn't be to win, but rather to learn and gain valuable experience. Experiencing that feeling of presenting your solution to industry experts is probably much more critical for your personal development than winning. The part of the hackathon where the most learning takes place is during the questions after your presentation. Most of the judges are industry experts and ask questions that really make you think about your solution. They also often suggest ways to improve your answer that comes in handy the next time you need to perform a similar task.

I would advise students to get a few friends together and see it as an opportunity to have some fun, drink some coffee and get out of their comfort zone. As students, we can get so caught up in getting good marks that we almost lose sight of how important it is to apply what we are learning. Hackathons provide an opportunity to test yourself in an environment where you have nothing to lose.