Adaptive Fraud Detection Using Benford’s Law

  • Fletcher Lu
  • J. Efrim Boritz
  • Dominic Covvey
Part of the Lecture Notes in Computer Science book series (LNCS, volume 4013)


Adaptive Benford’s Law [1] is a digital analysis technique that specifies the probabilistic distribution of digits for many commonly occurring phenomena, even for incomplete data records. We combine this digital analysis technique with a reinforcement learning technique to create a new fraud discovery approach. When applied to records of naturally occurring phenomena, our adaptive fraud detection method uses deviations from the expected Benford’s Law distributions as an indicators of anomalous behaviour that are strong indicators of fraud. Through the exploration component of our reinforcement learning method we search for the underlying attributes producing the anomalous behaviour. In a blind test of our approach, using real health and auto insurance data, our Adaptive Fraud Detection method successfully identified actual fraudsters among the test data.


Optimal Policy Outlier Detection Blind Test Fraud Detection Digit Frequency 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Lu, F., Boritz, J.E.: Detecting Fraud in Health Insurance Data: Learning to Model Incomplete Benford’s Law Distributions. In: 16th European Conference on Machine Learning, Porto, Portugal, pp. 633–640. Springer, Heidelberg (2005)Google Scholar
  2. 2.
    Bolton, R.J., Hand, D.J.: Statistical Fraud Detection: A Review. Statistical Science 17(3), 235–255 (1999)MathSciNetGoogle Scholar
  3. 3.
    Nigrini, M.J.: Digital Analysis Using Benford’s Law. Global Audit Publications, Vancouver, B.C., Canada (2000)Google Scholar
  4. 4.
    Nigrini, M.J.: Can Benford’s Law Be Used In Forensic Accounting? In: The Balance Sheet, June 7–8, 1993 (1993)Google Scholar
  5. 5.
    Pinkham, R.S.: On the Distribution of First Significant Digits. Annals of Mathematical Statistics 32, 1223–1230 (1961)zbMATHCrossRefMathSciNetGoogle Scholar
  6. 6.
    Hill, T.P.: A Statistical Derivation of the Significant-Digit Law. Statistical Science 4, 354–363 (1996)Google Scholar
  7. 7.
    Carslaw, C.A.P.N.: Anomalies in Income Numbers: Evidence of Goal Oriented Behaviour. The Accounting Review 63, 321–327 (1988)Google Scholar
  8. 8.
    Crowder, N.: Fraud Detection Techniques. Internal Auditor (April 17–20, 1997)Google Scholar
  9. 9.
    Sutton, R.S., Barto, A.G.: Reinforcement Learning: An Introduction. MIT Press, Cambridge (1998)Google Scholar
  10. 10.
    Sutton, R.S.: Learning to predict by the method of Temporal Differences. In: Machine Learning, vol. 3, pp. 9–44 (1988)Google Scholar
  11. 11.
    Lu, F., Schuurmans, D.: Monte Carlo Matrix Inversion Policy Evaluation. In: UAI: Proceedings of the 19th Conference, pp. 386–393. Morgan Kaufmann, San Francisco (2003)Google Scholar
  12. 12.
    Nigrini, M.J., Mittermaier, L.J.: The Use of Benford’s Law as an Aid in Analytical Procedures. Auditing: A Journal of Practice and Theory 16(2), 52–67 (1997)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2006

Authors and Affiliations

  • Fletcher Lu
    • 1
  • J. Efrim Boritz
    • 2
  • Dominic Covvey
    • 2
  1. 1.Canadian Institute of Chartered AccountantsScarborough, Ontario
  2. 2.University of WaterlooWaterloo, OntarioCanada

Personalised recommendations