Author: David Sragli, BSc Physics, University of Aberdeen
Project: Using machine learning to reduce localization error in auto-telemetry systems
Supervisor: Justin Travis, School of Biological Sciences, University of Aberdeen
I’m a BSc Physics student at the University of Aberdeen, entering third year this term. I had the opportunity over the Summer 2022 to participate in a research project “Using machine learning to reduce localization error in auto-telemetry systems” as an intern.
I have a huge interest in machine learning and even though I am Physics undergraduate I always had a fascination towards Biology. Hence, this project was close to my heart, since I was able to utilize my machine learning skills and learn new things about an area that I am not that familiar with.
During the project I joined a research group at the University of Aberdeen, where I worked on possible ways of reducing the localization error in auto-telemetry systems. We explored more than 5 machine learning approaches and we achieved significant results.
The project was mostly led by a PhD student and a Professor at the University of Aberdeen. My day-to-day experience varied during the time with the project. At the beginning I was mostly reading the relevant research papers, so I would be up to date with the current findings. After the reading was finished, we started analyzing the data that was recorded in Columbia. Drone flight data and walking data was recorded that mimicked the movement of birds, frogs or other animals that would be observed and studied in the grid system later on. Based on the data we built Machine learning models that would help us reduce the localization errors, that would happen because of the noise or that the signal is not travelling through vacuum. We used Python programming language along with my python frameworks to tackle this task of reducing the errors.
I have found this opportunity to be extremely rewarding. I was able to delve deeper into Machine learning and learnt new things how this problem could be solved. Furthermore, the experience of being in a research group was an even more valuable experience. The research group supervisors were super helpful, and they really made this placement to be as good as it can be. They also opened my eyes up to how a papers are established and how they are published, which are going to be cherished skills that I will surely be using in the future.
Photo by Niels van Altena on Unsplash