Java Reference
In-Depth Information
Many players in the retail industry already leverage data mining
extensively. Customer relationship management (CRM) and the
desire for one-to-one marketing [Peppers/Rogers 1999] make good
use of various cross-industry solutions: customer acquisition and
retention, response modeling, and new product line development,
among others. Loyalty programs such as those providing affinity
cards (frequent buyer cards) allow retailers to understand the buying
habits of customers and predict future behavior and needs.
Retail problems, however, go beyond CRM. Efficiently processing
and managing inventory can make a significant difference in profit
margins. Wu [2002] states,
In retail, every time merchandise is handled it costs the merchant. By incorporating
data mining techniques, retailers can improve their inventory logistics and thereby
reduce their cost in handling inventory. Through data mining, a retailer can identify
the demographics of its customers such as gender, martial status, number of children,
etc. and the products that they buy. This information can be extremely beneficial in
stocking merchandise in new store locations as well as identifying 'hot' selling
products in one demographic market that should also be displayed in stores with sim-
ilar demographic characteristics. For nationwide retailers, this information can have
a tremendous positive impact on their operations by decreasing inventory movement
as well as placing inventory in locations where it is likely to sell.
Life Sciences
Life sciences typically involves research that analyzes the structure,
function, growth, origin, evolution, or distribution of living organisms
and their relations to their natural environments [NCCS 2005]. It is a
fruitful area for applying data mining techniques amidst the deluge
of data confronting the life sciences industry [Lanfear 2006]. Problems
cover a wide range of areas including disease diagnosis [Chan
2002] [Lerner
2001] and treatment, genomics, drug interactions,
drug discovery, and cancer research [May/Heebner 2003].
For assessing disease treatments [Hamm 2004], attribute importance
techniques can rank treatments, treatment factors, and treatment effica-
cies. For example, factors associated with positive diabetes treatments
are ordered based on the drug received, full patient history, number of
hospital admissions, gender, etc. In addition, association techniques can
identify correlations between a particular treatment and patient out-
come. Rules may be of the form, “if number of visits to provider > 5
then outcome
improvement in 36 percent of cases.”
Search WWH ::

Custom Search