The subconscious mined : beneath the surface of a myriad of business data, meaningful patterns can be found by the sophisticated data-mining techniques of neutral networks
Date
2008-02
Authors
Olivier, Riaan
Gevers, Wim
Journal Title
Journal ISSN
Volume Title
Publisher
Stellenbosch : University of Stellenbosch Business School
Abstract
The concept of self-organising feature
maps (SOFMs) is a powerful tool that
can be used in a variety of applications
for data-mining and exploration,
as well as decision support. In its
basic form, an SOFM reduces the parameters
that describe a specific application, making it
easier for users to understand the underlying
problem. By enabling the visualisation of the
reduced parameters, it adds the strength of
human visual interpretation to complex
decision-making scenarios.
Applications of these tools may be found in
bankruptcy prediction, control of industrial
processes and in a variety of other instances
with time-series data, as was the case in a study
conducted at the University of Stellenbosch
Business School (USB). This research examined
how SOFMs could help solve the logistical
planning of a complex product.
Description
Keywords
Self-organising feature maps, SOFMs, Logistical planning, New products -- Planning