ZeroTune ICDE 2024
Published:
I had the incredible opportunity to present our paper, โZeroTune: Learned Zero-Shot Cost Models for Parallelism Tuning in Stream Processing,โ at an A* venue IEEE ICDE Conference 2024 ๐
Our research focuses on developing novel zero-shot cost models for performance prediction and efficient parallelism tuning in stream processing systems. ZeroTune aims to accurately predict performance and generalize across unseen parallel query structures and heterogeneous resources, all while reducing computational costs and training effort through a data-efficient training strategy. This represents a significant advancement in the field of data stream processing, and Iโm thrilled to have shared our findings with such an esteemed audience.
๐ See the full update on LinkedIn
๐ Our paper:
