Description Usage Arguments Value References Examples
View source: R/GetDetectorScore_normalizeScore.R
ReduceAnomalies
It reduces the number of detected anomalies. This function is
designed to reduce the number of false positives keeping only the first detection of all those
that are close to each other. This proximity distance is defined by a window
1 | NormalizeScore(real.score, perfect.score, null.score)
|
real.score |
Detector score. See |
perfect.score |
Perfect detector score; one that outputs all true positives and no false
positives. See |
null.score |
Perfect detector score; one that outputs all true positives and no false
positives. See |
Normalized score.
A. Lavin and S. Ahmad, “Evaluating Real-time Anomaly Detection Algorithms – the Numenta Anomaly Benchmark,” in 14th International Conference on Machine Learning and Applications (IEEE ICMLA’15), 2015.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 | ## Generate data
set.seed(100)
n <- 180
x <- sample(1:100, n, replace = TRUE)
x[70:90] <- sample(110:115, 21, replace = TRUE)
x[25] <- 200
x[150] <- 170
df <- data.frame(timestamp = 1:n, value = x)
# Add is.real.anomaly column
df$is.real.anomaly <- 0
df[c(25,80,150), "is.real.anomaly"] <- 1
## Calculate anomalies
result <- CpSdEwma(
data = df$value,
n.train = 5,
threshold = 0.01,
l = 3
)
res <- cbind(df, result)
# Get null and perfect scores
np.scores <- GetNullAndPerfectScores(df)
np.standard <- np.scores[1,]
np.fp <- np.scores[2,]
np.fn <- np.scores[3,]
# Get detector score
scores <- GetDetectorScore(res, print = FALSE, title = "")
# Normalize standard score
NormalizeScore(scores$standard, np.standard$perfect.score, np.standard$null.score)
# Normalize low_FP_rate score
NormalizeScore(scores$low_FP_rate, np.fp$perfect.score, np.fp$null.score)
# Normalize low_FN_rate score
NormalizeScore(scores$low_FN_rate, np.fn$perfect.score, np.fn$null.score)
|
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.