Implements Danai Koutra's TensorSplat algorithm
tensorsplat.RdImplements Danai Koutra's TensorSplat algorithm
Arguments
- matlist
The list of matrices where each matrix denotes the adjacency matrix of the network.
- k
The number of components in PARFAC tensor decomposition.
- alpha
The threshold to declare anomalies
References
Koutra, D., Papalexakis, E. E., & Faloutsos, C. (2012). TensorSplat: Spotting latent anomalies in time. Proceedings of the 2012 16th Panhellenic Conference on Informatics, PCI 2012, 144–149. https://doi.org/10.1109/PCi.2012.60
Examples
# We generate a series of networks and add an anomaly at 50th network.
set.seed(1)
networks <- list()
p.or.m.seq <- rep(0.05, 50)
p.or.m.seq[20] <- 0.2 # anomalous network at 20
for(i in 1:100){
gr <- igraph::erdos.renyi.game(100, p.or.m = p.or.m.seq[i])
networks[[i]] <- igraph::as_adjacency_matrix(gr)
}
tensobj <- tensorsplat(networks, k = 2)
#>
|
| | 0%
|
|=== | 4%
|
|====== | 8%
|
|======== | 12%
|
|=========== | 16%
|
|============== | 20%
|
|================= | 24%
|
|==================== | 28%
|
|====================== | 32%
|
|========================= | 36%
|
|============================ | 40%
|
|=============================== | 44%
|
|================================== | 48%
|
|==================================== | 52%
|
|======================================= | 56%
|
|======================================================================| 100%
tensobj # anomalous networks
#> [1] 20 30 23 32 13