We are thrilled to announce that Lamarr Institute researcher Fouad Alkhoury has been honored with the Best Student Paper Award at this year’s Discovery Science 24 conference. Together with co-author and Lamarr Alumnus Dr. Pascal Welke from TU Wien, Fouad presented the groundbreaking work titled “Splitting Stump Forests: Compressing Large Random Forest Models for Edge Devices”.
The awarded research introduces an innovative technique called Splitting Stump Forests, which compresses large random forest models into a compact form suitable for memory- and resource-constrained edge devices. By selecting key decision nodes and combining supervised methods with unsupervised balance measures, the approach maintains the predictive performance of the original model while significantly reducing its size. This advancement makes AI applications more adaptable to real-world limitations and is a promising step towards efficient and high-performance AI solutions.
Reflecting on the motivation behind this research, Fouad said, “The goal is to make AI more efficient and practical for real-world applications, especially on resource-constrained devices. This aligns closely with Lamarr Institute’s mission to drive innovation through high-performance, efficient AI models while addressing real-world limitations.”
Winning the Best Student Paper Award at Discovery Science 24 is a significant achievement, as this conference is known for its focus on advancing scientific knowledge discovery and fostering intensive discussions on innovative approaches to intelligent data analysis. This recognition highlights the relevance and impact of Fouad’s research within this dedicated community.
Explore the Research Further
For those interested in a deeper dive into this award-winning work, the full paper will soon be published in the Springer proceedings. In the meantime, you can access it here.
Congratulations to Fouad Alkhoury and Dr. Pascal Welke on this outstanding achievement!
