Complimentary from the publishers of Predictive Modeling News August 1, 2012 | ||
Complimentary Webinar: How much is your data costing you? Implications of outdated, incomplete and missing health plan data. Register Today. |
||
"The term ‘big data’
has been largely used to describe the vast amounts of Internet browsing
tracking data becoming more available. In healthcare, AHRQ’s HCUP data
set is an example of ‘big data,’ as are other publically available
healthcare claims data sets, such as New York’s SPARCS. Basically, any
large data set would be primarily used for three things: finding
patterns, discovering unusual correlations and mining data. As Health
Information Exchanges develop, each one will have large amounts of data,
and all of the HIEs nationally combined certainly would qualify as a
vast store of data on how care is delivered. From a predictive analytics
standpoint, decision-tree methods would probably be especially relevant,
since their objective is to construct groups of similar records based on
the discovery of shared attributes that might not be readily apparent.”
Lawrence S. Borok, President, Vantage Point Healthcare Information
Systems |
||
In a recent FICO survey, 64% of insurers said they lacked the ability to
rapidly deploy or update models to maximize business impact. One-third
said that implementing a new model takes four to six months, and 51%
said it takes six months or longer. Excerpted from Predictive Modeling News, Volume 5, Number 8, FICO Solution Accelerates ROI for Insurance Analytics |
||
Analytics Author: New ‘Decision Frontiers’ are Coming Health Data Management, July 26, 2012 |
||
Roadmap: SAP BusinessObjects SAP, July 17, 2012 |
||
Health Care Fraud: Analytics is the RX for $250 Billion in Fraudulent
Health Care Costs Huffington Post, July 17, 20122 |
||
Solving Our Diabetes Problem With A System That Predicts Who Needs Help Co.Exist, July 16, 2012 |
||
Predictive Modeling Is a Crucial Tool for Fighting Health Care Fraud,
Association Says Bureau of National Affairs, July 13, 2012 |
||
|
||
There's no cost to participate in the Predictive Modeling News LinkedIn Group where you can Network, exchange information, and follow current developments with other professionals interested in predictive modeling. You can sign up now at: http://www.linkedin.com/groups?gid=1669467 | ||
Predictive Modeling Bulletin is a complimentary publication issued periodically from the Health Policy Publishing, LLC, the publisher of Predictive Modeling News (www.predictivemodelingnews.com) Your friends and colleagues can subscribe to Predictive Modeling Bulletin at no charge by visiting: http://www.predictivemodelingnews.com/resources.html | ||