Bayesian Optimization of Gaussian Processes with Applications to Performance Tuning
summary
Introduces Bayesian optimization using Gaussian process models (Gaussian Processes, Bayesian Optimization) and explains why it's suitable for efficiently exploring high-dimensional configuration spaces. Describes Autotune, a service implemented at Twitter that applies Bayesian optimization to tune performance of a modern microservices architecture, covers engineering challenges encountered during implementation, and presents examples of performance improvements achieved in data-center operation.
AI-written summary · about
chapters (13)
- 0:00Intro
- 0:11Bayesian Optimization of Gaussian Processes applied to Performance Tuning
- 1:56A JVM Engineer talks to a Data Scientist
- 9:55Mining for Gold
- 14:40Applications
- 19:23Black Box Modeling
- 20:28Model Parameters
- 23:59Probabilistic Models
- 34:26Covariance Kernel Functions
- 36:19Thompson Sampling
- 37:04Probability of Improvement
- 37:45Upper Confidence Bound
- 38:16Maximizing the Acquisition Function
related talks
28mBayesian Optimization- Fundamentals, Implementation, and Practice
24mUsing Bayesian Optimization to Tune Machine Learning Models
34mBayesian Optimization and it's application to Neural Networks
39mBayesian optimisation with scikit-learn
46mBayesian Optimization Accelerator
41mPlanes, Trains, and Skateboard Shoes - Bayesian methods in engineering and product design
41mOptimizing Machine Learning Models
41mAutomated Performance Tuning with Bayesian Optimization