An implementation of sparse Gaussian Markov random field mixtures presented by Ide et al. (2016) <doi:10.1109/ICDM.2016.0119>. It provides a novel anomaly detection method for multivariate noisy sensor data. It can automatically handle multiple operational modes. And it can also compute variable-wise anomaly scores.
| Version: | 0.3.0 |
| Imports: | ggplot2, glasso, mvtnorm, stats, tidyr, utils, zoo |
| Suggests: | dplyr, ModelMetrics, testthat, covr, knitr, rmarkdown |
| Published: | 2018-04-16 |
| Author: | Koji Makiyama [cre, aut] |
| Maintainer: | Koji Makiyama <hoxo.smile at gmail.com> |
| License: | MIT + file LICENSE |
| NeedsCompilation: | no |
| Materials: | NEWS |
| CRAN checks: | sGMRFmix results |
| Reference manual: | sGMRFmix.pdf |
| Vignettes: |
Sparse Gaussian MRF Mixtures for Anomaly Detection |
| Package source: | sGMRFmix_0.3.0.tar.gz |
| Windows binaries: | r-devel: sGMRFmix_0.3.0.zip, r-release: sGMRFmix_0.3.0.zip, r-oldrel: sGMRFmix_0.3.0.zip |
| macOS binaries: | r-release: sGMRFmix_0.3.0.tgz, r-oldrel: sGMRFmix_0.3.0.tgz |
| Old sources: | sGMRFmix archive |
Please use the canonical form https://CRAN.R-project.org/package=sGMRFmix to link to this page.