A
Demo on Knowledge Discovery
I have prepared
a demo about some features of Rough Sets Theory and Formal Concept Analysis
applied to Data Mining and Knowledge Discovery in Databases.
Drop
me a line and you will receive it (minimal requirements: Windows and
1.5M free).
This demo was presented
at the University of Tokyo, during the Fourth International Workshop
on Rough Sets, Fuzzy Sets, and Machine Discovery.
The demo features
data mining and knowledge discovery functions related to a database representing
72 socio-economic European sub-regions evaluated by 12 socio-economic parameters.
Using an interactive
map of Europe and a control panel, you will be able to:
-
display the evaluation
of any sub-region and the meaning of any parameter;
-
group the sub-regions
by means of their evaluations;
-
group the sub-regions
by means of user selected parameters and user selected groups of evaluations;
-
check your own
hypotheses about the possible dependence between two sets of parameters;
-
make decisions
about a user selected set of sub-regions depending on a user selected set
of parameters (any selected sub-region and any selected parameter may be
weighed): the machine will check the consistency and the reliability of
your decision, offering solutions and computing minimal sets of parameters
on which your decision actually depends;
-
chose a set of
sub-regions by clicking on the map: the machine will suggest other sub-regions
explaining why they are related to the sub-regions that you have selected.
-
..............................
Any function has an
hypertext interface, so that you can move freely, for example, from a result
about a set of decisions to the explanation of the meaning of the minimal
set of parameters required by that decision and from this last window select
all the sub-regions with a particular evaluation. From the resulting group
you can then select a sub-region and display its evaluation, without
getting lost: your navigation will be always traced.
Home