Statistical Methods for Crime Series Linkage
Statistical Methods for Crime Series Linkage. This package provides
code for criminal case linkage, crime series identification, crime series
clustering, and suspect identification.
crimelinkage 0.0.4 (09/18/2015)
- Switched to the formula interface for naiveBayes() to be more consistent with other R functions.
- There is also a naiveBayes.fit() function to pass in data w/o formula.
- Note: this is different than the function from previous version. Old calls
will no longer work. Please update to the new interface.
- Added S3 methods: plot.naiveBayes, predict.naiveBayes
- Improved compareCrimes() function.
- Better estimation of the expected absolute differences between the temporal
crime variables using convolution. It is now a deterministic function.
- Uncensored event times, if exact times are known, are now permitted.
- Updated crimes dataset with a slight modification to some dates.
- Updated predict.naiveBayes() to give warning when a predictor is not included
in model, but continue to give results.
- Corrected typos in vignettes
- Added default packges to imports
crimelinkage 0.0.3 (04/18/2015)
- Initial package submission