Information Technology Reference
In-Depth Information
We measure the predication results in terms of the following two factors:
completeness, that is the rate of predication fishing grounds over the actual ones;
precession refers to the difference between prediction fishing grounds and
actual fishing grounds. The fishery experts thinks that the precision is up to
80% once the difference is less than a fishing district. The experiments indicate
the average prediction precision of the system is up to 78%, and our system has
the value to be popularized(see Fig. 5.8). We have already finished the
demonstration test, and now are cooperating with Shanghai Fishery Politics
Bureau to apply the system to production.
Exercises
1.
Explain what CBR is in your own words and describe several CBR
systems briefly.
2.
Describe the representation forms of cases and the basic thought
underlying CBR.
3.
Illustrate some applicable occasions for CBR ; and sketch the general
process of CBR in terms of the flow diagram
4.
What is the advantage of CBR over Rule-Based Learning (RBR)? Where are
the abilities of CBR
CBL
systems.
5.
List some kinds of similarities between the target cases and base ones; and
try to compare the characteristics of some mainly-used methods for
similarity measurement.
6.
Refer to relevant materials and introduce an application system based on
CBR; then discuss key challenges posed by CBR systems and how to
overcome them.
7.
Demonstrate the differences by examples between CBR systems and
expert systems.
