;
This web site and all of its contents Copyright © 2002 Peer Science LLC. All rights reserved
|
Fuzzy System Demonstrations
To see a demonstration of our
eXtreme Rules
TM intelligent data analysis component, select one
of the examples below.
Stock Market Strategies
Strategies for selecting stocks are
naturally expressed as rules. This demonstration illustrates the differences
between fuzzy logic and binary logic in expressing and processing rules. Two
parallel sets of rules are shown, one fuzzy, the other binary. Some of the
important properties illustrated are:
- Fuzzy rules are allowed to contradict, while crisp rules cannot.
- Information is lost in binary rule processing, as the result is either true
or false, while fuzzy logic preserves information, providing a result with a
continuous range of values.
- The eXtreme Rules component is a hybrid inference engine, processing rules
that may have both fuzzy and binary components.
Note: This is intended as a demonstration only! We are not advising you
on stock purchases. Our intent is only to illustrate characteristics of our
inference engine.
|
|
Case-Based Reasoning Demonstrations
To see a demonstration of our
eXtreme Cases
TM intelligent data analysis component, select one
of the examples below.
Bass Fishing Advisor
Fishing for largemouth bass is much more than putting a worm on a hook and waiting.
It has become a sophisticated combination of art and science that challenges clever minds. Hundreds of magazine
issues, dozens of books and hundreds of hours of television programming were used to build a rule-based system
for advising bass fishermen on their choice of lure and technique. But this alone is not enough. Rules need to be
combined with experience to account for local differences and preferences. This is an ideal application for
case-based reasoning. The BassExpert site demonstrates the power of combining a rule-based expert system for general
knowledge with a case-based reasoning system for adapting to local conditions. Some of the
important properties illustrated are:
- A rule-based system encodes general domain knowledge.
- As the user records his/her fishing experiences in an on-line fishing log, this data is
immediately available for adapting the advice to the particular fisherman.
- This hybrid system is readily adapted to new domains where general rules can be formulated, and adaptation to
changing experience can be exploited. Domains particularly suitable for this technology are
- Call Centers
- Decision Support Systems
- Fraud Detection
- Customer Relationship Management
Note: The BassExpert application is a live commercial web site,
bassexpert.com.
|
|
|