TutorTeddy.com, Tutoring, Dallas, TX
Like us for free help*

Predictive Analytics Attains Strategic Objectives Across Business Functions. Objective - Enforce



Ask any Statistics/Probability/Math Question

The third reason for the need of predictive analytics is Enforce and this is done by maintaining business integrity which is done by managing fraud.


Social Network Analysis

Enforce - Maintain Business Integrity by Managing Fraud

For identification of suspicious claims many companies are looking forward to technology. Suspicious claims tells about when and how the claim is made, what the claim is, and who is involved in making claims. Here the motivation is to identify patterns and for assessing commonalities.


The connections are related to crime based fraud. There exist so many degrees of separations between the participants. Here is the situation where social network is helpful. With the help of social network analysis the previous unseen links can be uncovered. The tool -social network analysis tools assists in optimizing efforts for detecting fraudulent claims. The fraud scoring engine involves a combination of various analytical techniques. The analytical techniques includes:


  • Database Searches

  • Business Rules

  • Anomaly and Exception Reporting

  • Text Mining

  • Analysis of Network link

  • And Most Importantly Predictive Modeling

  • With the help of the above techniques the likelihood of a fraudulent claim can be determined.

Predictive Modeling

In detecting fraudulence in insurance the approach that is involved in predictive modeling is to analyze the set of known suspicious claims. It is believed that similar claims in the future can be identified from the existing data set. Although the technique is very effective and powerful it involves a large set of known suspected claims which is used for building model and also training people. Limited known suspicious claims fails to detect fraudulence , this is because with limited known suspicious claim one cannot proceed with a technology to detect fraudulence. But even if a few suspicious claims is identified in past, a solid predictive analytic solution can be build by considering numerous statistical approaches.


For making an approach to hybrid fraud detection text analytics works wonder. A seasonal investigator wants to read the claim notes for the same reason. Unstructured text data should be used in a predictive model. By means of advanced text mining tools variable and entity are extracted. In some cases the data elements that are used in predictive analytics for fraud detection are collected from unstructured data sources.


Text Analytics

TutorTeddy.com & Boston Predictive Analytics

[ Email your Statistics or Math problems to tutor@aafter.com (camera phone photos are OK) ]


Boston Office (Near MIT/Kendall 'T'):
Cambridge Innovation Center,
One Broadway, 14th Floor,
Cambridge, MA 02142,
Phone: 617-395-8864


Dallas Office (Near Galleria):
15950 Dallas Parkway,
Suite 400,
Dallas, TX 75248,
Phone: 866-930-6363

Copyright © 2011 tutorteddy.com. All Rights Reserved. | By using our site, you agree to our TOS.