Skip to content

Density estimator with density-change detection using wavelets.

Notifications You must be signed in to change notification settings

gednyengs/Estimator-Detector

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

8 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Estimator-Detector

Density estimator with density-change detection using wavelets.

##How to use the density change detector This 1D version of the density change detector comes bundled with the Matlab 1D density estimator. The change detector is enabled by default.

To disable the change detector:

  • open the changeDetectorInit.m file
  • Inside the initialization function, set CD.ENABLE to 0. The change detector is disabled when the value of CD.ENABLE is set to 0 and it is enabled when the value of CD.ENABLE is set to 1.

The current implementation of the change detector uses pairs of sliding windows and continuously compares the enclosed approximated density function by the two windows in the pair. If the two density functions differ significantly, the detector flags a change. It is important to note that the measure of similarity or dissimilarity between the density functions approximated by the two given windows in the pair depends on the distance measure used. With the current version of the change detector, we provide 3 distance measures: ksi, phi, and KS distances.

The number of pairs of windows to use can be set by changing the CD.NumOfPairs field within the change detector's initialization function. After defining the number of pairs of windows to use, it is essential to also set the sizes for the window pairs. The windows in the same pair get the same size. So the CD.WindowSizes field defines the window sizes for each pair of windows. Equally important is the need to set the threshold distances for the change detector's flag. Please set a threshold distance per pair of windows by changing the CD.Thresholds field.

If one is interested in detecting changes in particular regions of the domain of the density function, these regions can be defined in the CD.Regions field. for each region, please specify the two endpoints for the region as a row entry in the CD.Regions field with the first entry being the minimum value and the second entry being the maximum value.

About

Density estimator with density-change detection using wavelets.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published