Dr Zdravko Botev
The University of New South Wales
We speak to Dr Zdravko Botev who explains why mathematics is the language of the cosmos and the challenges researchers face in the machine learning environment
Why did you become a mathematician?
Mathematics is the only knowledge which can claim absolute truths with the certainty craved by religion. Ever since ancient times our understanding of the world has kept changing – we had to change our views about the origin of the universe; we had to change our beliefs about what makes us ill and how we can cure diseases. Yet in all this whirlpool of constant change, only mathematics gives us absolute security and certainty. What the ancient Greeks proved more than 2000 years ago is still as valid as when it was discovered, and will be valid for all eternity. It is often said that if one day an alien intelligence visits our Solar System, then mathematics will be our only means of communication. Given all this, how can one not desire to study the only universal language of the cosmos?
What drives your interest in this field?
My focus is on using computer simulated random numbers to mimic processes that exhibit randomness. Also, I design algorithms for the accurate and efficient computation of the probability of occurrence of rare and high-impact events. The area is called Monte Carlo simulation. Originally, the method was proposed to help carry out difficult computations necessary for the construction of the atomic bomb during WWII. These days Monte Carlo method have become widespread with many applications in data mining.
Do you have any advice for future researchers?
Pick the topic that you find interesting, not necessarily the topic that is hot at the moment.
What are the most interesting “big questions” or challenges facing researchers in your area?
The biggest challenge is to create computer-driven systems that act and learn intelligently and autonomously, in a way similar to humans. Nobody knows how to create such computers. At the moment “artificial intelligence” and “deep-learning” are simply a buzzwords for old-fashioned “curve-fitting”.
Dr Botev is delivering the Mathematical Methods for Machine Learning topic at the 2019 Summer School.