Extending research into evolutionary #AutoML, we present an approach that evolves algorithms from scratch—using only basic mathematical operations—rediscovering fundamental ML techniques & showing the potential to discover novel algorithms. Read more ↓ https://t.co/XtNsC9qFnj
To support research into cloud computing and cluster scheduling, we are releasing a new 29-day trace of the Borg cluster management system, which captures scheduling decisions and resource usage for Google’s computational fleet. Learn more & get the data ↓https://t.co/KI7jYYEfm3
Our paper on learning data layouts in storage for big data analytics will appear at #SIGMOD2020. We propose a data structure called a qd-tree (built using deep RL) to layout data blocks in a workload-guided manner. Easy to integrate with DB as well! Paper: https://t.co/Qy0VF5J63P