frac: FRaK: Feature Modeling Approach to Anomaly Detection
In Stevo15025/FRaC: Anomaly Detection

Description Usage Arguments Details Value Author(s) References

The FRaC Algorithm

1	frac(x, models, n.cv=2, keys = NULL grid=NULL, allowParallel = FALSE, tuneList = NULL)

`x`	An NxT matrix with T observations of N varialbles
`models`	a string vector of models available in the caret package
`keys`	A vector of integers representing places that the data set has time variables. This is only here for short term bug fixes
`n.cv`	An integer specifying the number of cross-validations
`allowParallel`	Logical definining whether or not the user would like to use a parallel backend if one is set up
`tuneList`	A list of models for train functions
`grid`	A list containing the different parameter values for each models iterations. Can be set to default NULL

Version 0.0.3 Alpha

values or sup: the normalised suprisal score for each observation. Higher scores equate to a higher chance of an observation being an outlier

Steve Bronder

K. Noto, C. E. Brodley, and D. Slonim. FRaC: A Feature-Modeling Appraoch for Semi-Supervised and Unsupervised Anomaly Detection. Data Mining and Knowledge Discovery, 25(1), pp.109—133, 2011.

Stevo15025/FRaC documentation built on May 9, 2019, 3:08 p.m.