Approximate counting: a detailed analysis
BIT - Ellis Horwood series in artificial intelligence
Probabilistic counting algorithms for data base applications
Journal of Computer and System Sciences
A linear-time probabilistic counting algorithm for database applications
ACM Transactions on Database Systems (TODS)
Space/time trade-offs in hash coding with allowable errors
Communications of the ACM
Dynamic Programming and Optimal Control
Dynamic Programming and Optimal Control
Maintaining Stream Statistics over Sliding Windows
SIAM Journal on Computing
New directions in traffic measurement and accounting
Proceedings of the 2002 conference on Applications, technologies, architectures, and protocols for computer communications
| Hi-index | 0.00 |
One considers the problem of estimating the call destination dispersion on telecommunications usage to use in fraud detection. The problem is that such detection needs to be performed for each individual customer and kept up to date at all times. The use of fast and small footprint algorithms is critical due to the huge number of events and customers to verify and since approximate answers are enough in most situations. This paper presents telecommunications customer behavior to justify the use of approximate estimators and then presents multiple options of algorithms to solve the problem. These algorithms present a novel approach to the moving window dispersion problem by the use of a probabilistic time decay mechanism.