Opened 9 years ago
#2498 new defect
Improve Gaussian process training and inference to support larger datasets
| Reported by: | gkronber | Owned by: | |
|---|---|---|---|
| Priority: | medium | Milestone: | HeuristicLab 4.0 |
| Component: | Algorithms.DataAnalysis | Version: | 3.3.12 |
| Keywords: | Cc: |
Description
The most recent and most relevant seems to be KISS-GP which promises big improvements especially for time-series or tempo-spatial data sets. See Andrew Gordon Wilson and Hannes Nickisch, "Kernel interpolation for scalable structured Gaussian processes (KISS-GP)", International Conference on Machine Learning (ICML), 2015 (http://www.cs.cmu.edu/~andrewgw/)
Note: See
TracTickets for help on using
tickets.
