Abstract | ||
---|---|---|
For large-scale distributed systems, the knowledge component at the core of the MAPE-K loop remains elusive. In the context of end-to-end probing, fault monitoring can be re-casted as an inference problem in the space-time domain. We propose and evaluate Sequential Matrix Factorization (SMF), a fully spatio-temporal method that exploits both the recent advances in matrix factorization for the spatial information and a new heuristics based on historical information. Adaptivity operates at two levels: algorithmically, as the exploration/exploitation trade off is controlled by a self-calibrating parameter, and at the policy level, as active learning is required for the most challenging cases of a real-world dataset. |
Year | DOI | Venue |
---|---|---|
2014 | 10.1109/ICCAC.2014.17 | ICCAC |
Keywords | Field | DocType |
historical information,learning (artificial intelligence),self-calibrating parameter,spatial information,smf,fault tolerant computing,fault prediction,exploration-exploitation tradeoff,sequential fault monitoring,active learning,fault diagnosis,sequential matrix factorization,matrix decomposition,matrix factorization,fully spatio-temporal method,distributed processing,prediction algorithms,estimation,yttrium,linear programming | Spatial analysis,Data mining,Active learning,Computer science,Inference,Matrix decomposition,Theoretical computer science,Exploit,Heuristics,Prediction algorithms,Linear programming | Conference |
Citations | PageRank | References |
0 | 0.34 | 22 |
Authors | ||
3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Dawei Feng | 1 | 11 | 2.57 |
Cécile Germain | 2 | 122 | 17.50 |
Julien Nauroy | 3 | 14 | 2.03 |