Quantitative and Qualitative analysis, Sabermetrics and Statistical Trends and Regression Analysis.
Analysis included creation/expansion of a basic regression for a large population as provided by the following:
∑Q1 revenue2007-2009 /3= Δr1; ∑Q2-4revenue2007-2009/3= Δr2-4, r=revenue for each month/n months A= Δr Q12010 B=Δr q2-42010
(x+a)^n=∑_(k=0)^n〖(n¦k) x^k a^(n-k) 〗
In 2013 and 2014 I created an algorithm for the BP oil spill compensation team characterized by the hybridization of a Bayesian multi-variate non-linear regression expressed as a two-tailed non-centralized hypergeometric distribution in a closed system without replacement. Algorithms were created to identify relationships and patterns in unstructured data.
This analysis was used to detect organized fraud, discovered the Watts & Guerra scheme and uncovered over $2.3 billion in fraudulent claims filed against the BP Settlement. Findings were submitted to the Freeh Group and other agencies.
Additional prior projects include large-scale: Anti-Money Laundering, Securities, Probate Fraud and Insurance Fraud reviews and analytics.
Overlapping qualitative data like that contained in the above email can be translated into metrics for quantitative analysis
Cross Referenced materials provided the foundation for the BP analytics to detect fraud.
First Generation analytics were derived from partial reports
Initial metrics were created for analysis
Indicia of a much larger scheme involving 'ghost' (non-existent) claimants became evident even from this first, primitive version of fraud analytics.
3+/4 ILR Spanish Language Proficiency
1/1+ ILR Russian Language Proficiency
Langely Joint Base National Honor Guard (Airforce)