By Michael E. Schuckers
Biometrics, the technological know-how of utilizing actual qualities to spot contributors, is enjoying an expanding position in our security-conscious society and around the globe. Biometric authentication, or bioauthentication, platforms are getting used to safe every little thing from enjoyment parks to financial institution money owed to army installations. but advancements during this box haven't been matched through an similar development within the statistical equipment for comparing those systems.
Compensating for this want, this precise text/reference offers a simple statistical method for practitioners and testers of bioauthentication units, delivering a suite of rigorous statistical equipment for comparing biometric authentication platforms. This framework of tools could be prolonged and generalized for quite a lot of purposes and tests.
This is the 1st unmarried source on statistical equipment for estimation and comparability of the functionality of biometric authentication structures. The ebook specializes in six universal functionality metrics: for every metric, statistical equipment are derived for a unmarried method that comes with self assurance periods, speculation checks, pattern measurement calculations, energy calculations and prediction periods. those equipment also are prolonged to permit for the statistical comparability and assessment of a number of structures for either self sufficient and coupled data.
Topics and features:
- Provides a statistical method for the commonest biometric functionality metrics: failure to sign up (FTE), failure to obtain (FTA), fake non-match fee (FNMR), fake fit cost (FMR), and receiver working attribute (ROC) curves
- Presents tools for the comparability of 2 or extra biometric functionality metrics
- Introduces a brand new bootstrap method for FMR and ROC curve estimation
- Supplies greater than one hundred twenty examples, utilizing publicly on hand biometric information the place possible
- Discusses the addition of prediction durations to the bioauthentication statistical toolset
- Describes sample-size and tool calculations for FTE, FTA, FNMR and FMR
Researchers, managers and judgements makers wanting to check biometric platforms throughout various metrics will locate inside this reference a useful set of statistical instruments. Written for an upper-level undergraduate or master's point viewers with a quantitative history, readers also are anticipated to realize the subjects in a regular undergraduate information course.
Dr. Michael E. Schuckers is affiliate Professor of records at St. Lawrence collage, Canton, new york, and a member of the guts for identity know-how Research.