Alex Zhu has just completed an Honours year in Statistics, as part of his degree in Actuarial Studies and Advanced Mathematics.
He speaks to us about his intriguing Honours research project, the collaborative team work of the past year, what first ignited his love for maths and science, his involvement with MathSoc, some future plans for his research and himself - and more!
Read on to learn more about Alex...
How did you first get interested in maths, and what led you to select this particular double degree?
Ever since I was little, I always had an affinity towards solving riddles and puzzles, so naturally I was drawn to mathematics and science. After high school, I wanted to learn more mathematics and apply my problem-solving skills to the real world. I chose Actuarial Studies because I wanted to gain an exposure to commerce as well as mathematics.
What has your Honours year been like? Can you briefly describe your Honours Project?
My Honours year had been very rewarding and eye-opening. I was challenged by many problems that interested me and it opened many areas of statistics, algorithms and programming for me. Also, it was a chance to build invaluable relationships with my supervisors and peers who had given me immense support throughout the year.
I could not ask for a better experience to wrap up university. My Honours Project is about building an R package on the minicomputer Raspberry Pi that can connect to different sensors to collect and analyse data that enters as a continuous stream.
Your Honours supervisor was Pierre Lafaye De Micheaux. How did you first hear about A/Prof Lafaye De Micheaux's research, and what appealed to you about it? How was the experience of working within his team?
I was drawn to Pierre’s research from my initial intentions of learning some programming in my Honours Year. He told me about this project related to development with the Raspberry Pi, which I didn’t know too much about at the time.
Working within Pierre’s research team (Pavlo Mozharovskyi and Fabien Navarro) was very rewarding. I learnt a significant depth of R and C++ programming from Pavlo and a mix of useful streaming data analytics techniques from Fabien and Pierre. From working with my supervisors, I started to appreciate how much talent and hard work researchers need to complete an academic project.
How has this research shaped your views on mathematics or statistics?
The research on streaming data analysis has opened my perspectives on statistics and algorithms techniques used in these lesser explored data types. It has made me realise that the knowledge I have learnt from university has only scratched the surface of maths.
It has also made me realise the journey it takes for me to get to the end of my degree. Without the help of my family, my peers, and my supervisors, I would not have been able to appreciate the knowledge I have learnt from all the courses. At the end of the day, “Mathematics is about people” – Dr Diana Combe (paraphrased by [PhD student] Yudhi Bunjamin).
What is the relevance between Raspberry Pi and statisticians?
Statisticians can use the Raspberry Pi to collect low cost and easy to maintain real life data streams. This can potentially give rise to more research into techniques and algorithms in this area which is currently very computer science focused.
Data can be acquired more easily for scientists and statisticians, which can reduce the budget for research with environment data (Temperature, Sound, Infrared, Toxic Gas Level, etc.).
Are there any new statistical or machine learning theories, methodologies, and algorithms that need to be developed further to unleash the power of your new R package?
Yes! We believe there will be more opportunities to analyse data stream methods using statistics and estimation theory by providing this R package. The existing randomised algorithms opens a rich area to study, especially when we would like to calculate approximate quantities with a strict requirement on memory. One famous example is Monte Carlo Methods.
Do you plan to publish the results of your research, or to present them at a conference?
Yes - the plan will be to present the results to the R community, for example in the UseR! conference.
What courses did you select during your studies to arrive at your particular Honours research area?
I did the required courses for a Statistics major and Actuarial Degree. I also completed some extra Computer Science courses such as COMP2521 (Data Structures and Algorithms), COMP3121 (Algorithms and Programming Techniques) and COMP2041 (Software Constructions).
Have you been able to apply your maths and stats skills in other ways outside of your studies?
Definitely! I did several case competitions in my undergraduate years and I felt that machine learning and statistical learning theory were hot topics in business analytics. Many of the popular models used concepts from courses as early as First Year Linear Algebra and Calculus.
This is not limited to data analytics and actuarial. Financial Mathematics and Derivatives Pricing is also some of the careers that my friends have taken, and they did mention the usefulness of Stochastic Processes and Higher Analysis in these Quantitative areas of the financial markets.
You were one of MathSoc’s Executive Team members in 2020, looking after their Education portfolio. Can you tell us a little about your work with them?
Working with the UNSW Mathematics Society was a great opportunity to give back to the community and help younger students better adjust to university maths. It gave me a chance to teach and present first and second-year maths, which also helped my understanding.
At the tail-end of your Honours year, have you decided on your next steps, career or otherwise?
I am fortunate enough to have received a grad offer from the Actuarial Consulting Firm, Taylor Fry, next year. Thus, I will be joining them as an Actuarial Consultant to complete my training in my Actuarial pathway.
I am also considering doing a PhD, potentially in the future after finishing my Actuarial Qualifications.
Interview conducted by Susannah Waters