Coursera Flash Sale
40% Off Coursera Plus for 3 Months!
Grab it
Learn about Swift, a novel Bayesian Optimization approach for parameter configuration tuning in big data systems through this 18-minute conference presentation from USENIX ATC '25. Discover how researchers from Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Tsinghua University, and Huawei Cloud leverage generative adversarial networks (GANs) to generate high-quality configurations based on the highest-performing evaluated configurations. Explore how mixing these GAN-generated configurations with randomly generated ones skews the search space toward optimal configurations, resulting in faster convergence and reduced optimization time. Examine substantial experimental results demonstrating Swift's significant performance improvements over state-of-the-art approaches on Apache Flink, Spark programs, and industrial settings, all achieved in dramatically shorter timeframes. Gain insights into cutting-edge techniques for accelerating performance tuning in distributed data processing systems through the innovative application of generative AI methods.
