From kathryn.ross at state.or.us Tue Dec 14 08:30:05 2010 From: kathryn.ross at state.or.us (ROSS Kathryn * SCD SARS) Date: Tue, 14 Dec 2010 08:30:05 -0800 Subject: [CAFR-Contacts-News] FREE CPE Training: Understand how data mining can prioritize audits and investigations to ensure the most return on investment. Message-ID: <8A8CF693A3BB294DB354A999D49CF818077E7D9E@exchnode01.ad.state.or.us> This training event is free (sponsored by the State Controller's Division - SARS). This is an audio conference to be held at the Veteran's Building Auditorium. Space is limited to 100 people. Registration for this event will be coordinated through iLearn Oregon ( https://ilearn.oregon.gov ); additional information is listed in the "Registration" section of the below announcement. If you have questions, please contact Aaron Wallace at Aaron.Wallace at state.or.us or 503-373-7277, ext. 281. Thank you. ________________________________ Needles in a Haystack-Data Mining and Predictive Analytics to Prioritize Leads and Highlight Risk for Investigators and Auditors Jan. 12, 2011 - 2 CPE Hours Available! AGA is pleased to announce a new topic to our audio conference schedule-data mining for investigators and auditors. Government investigators and auditors are often overwhelmed by high volumes of data and can perform only ad-hoc analyses and queries to find areas of potential risk and fraud. This audio conference will reveal how data mining can prioritize audits and investigations on the highest value leads or risk areas to ensure the most return on investment. Data mining uses proven statistical techniques to identify anomalous transactions, contracts or people to rank the most likely fraudulent cases for investigation or audit. Predictive modeling employs statistical models to learn from past audits and investigations to predict, in real-time, which new cases are most likely to be improper or fraudulent. Searching for fraud is like looking for needles in a haystack, and specialized techniques are needed to succeed. When beginning a predictive modeling project, analysts almost always have relatively few known instances of fraud with which to represent the problem in the training phase of building the model. Some fraudulent cases that have escaped detection are mislabeled as valid, which adds to the challenge of distinguish between the two. Statistical outliers in the data, which are signs of unusual activity, are fruitful places to search for fraud, but a typical "alert" system built only on these will have far too many "false alarms," or false positives, to be useful in practice. In this audio conference, our presenters will show how to make real advances despite these challenges. To share their practical experience in data mining, audits and investigations are John Elder, Ph.D., CEO and founder of Elder Research, Inc., a thought leader in the field of data mining and predictive analytics, an award-winning book author and a frequent keynote speaker at conferences; John V. Kelly, CGFM, CPA, CFE, director of Forensic Audits, Office of Inspector General, U.S. Department of Homeland Security; and Edward Slevin, CISA, director of Computer-Aided Assessment Techniques, Office of Inspector General, U.S. Department of Education. Please join us for two hours of lively discussion about this important and timely topic. In addition to the speakers' commentary, about 20 minutes will be set aside for Q & A so that the participants can ask the speakers questions and share their own experiences. Date: Wed., Jan. 12, 2011 Time: 11-12:50 p.m. Pacific Standard Time Learning Objectives: To understand how data mining can prioritize audits and investigations on the highest value leads or risk areas to ensure the most return on investment. Prerequisite: Basic familiarity with data mining, risk, audit and investigations Advance Prep: None required Field of Study: Auditing CPE: Two credit hours Cost: FREE (Sponsored by SCD) Registration: https://ilearn.oregon.gov Registration for this event will be coordinated through iLearnOregon using the link provided above. You will need an active iLearnOregon account to complete the registration process. If you do not already have an iLearnOregon account, you can find information on how to create an account (for both state & non-state personnel) on the main log-in page for iLearnOregon. Once you have logged-in you will find a link for "How to Register For A Course" in the left navigation panel under "iLearnOregon Help Center." Sponsored by: -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: image/jpeg Size: 4777 bytes Desc: image001.jpg URL: