Lehrinhalte
During the initial phase of this class, students will hear presentations on how to read and review scientific papers, and how to give a good presentation. Additionally, all participants will receive a raw submission version of a top-tier conference research paper. Task of the students is to critically read the paper and prepare a written review of the paper, following a typical conference reviewing template. Afterwards, students will receive the actual expert reviews of the paper, compare these to their findings, and prepare and deliver a presentation that discusses the improvements made between the original submission and the publication of the paper based on the reviewers' criticisms. Moreover, each week, students will be required to reflect on the main challenges addressed in each of their peers' presentations. (Details will be announced in the class.)
Representative topics to be discussed, include data stream processing (e.g., scalable and parallel window joins, window aggregation, state management), data processing on modern hardware (e.g., GPU data processing, FPGA acceleration for data sketching), data processing for machine learning and data science (e.g., optimizing machine learning pipelines, dynamic parameter allocation in parameter servers), and query optimization and compilation (e.g., adaptive compilation).
