|
Gallery - Masa Takatsuka
JBean Studio - Visual Software Authoring System
Demand for effective analysis of:
-
massive amount of data
-
complex data
There is no best visualisation technique for a
given data set.
Interative Visualisation:
-
Interaction among different types of
visualisation techniques
-
customisable at the run-time
-
discovery of visual variable assignments.
GD-SOM - Geodesic Dome Self-Organisation
Map
-
The rapid technological improvement in the fields of electronics and computing
engineering
-
The increase in the amount of data and complexity of data.
-
More complex and sophisticated data analysis are demanded to solve challenging
problems
More
Sophisticated Analysis
-
Data mining using soft-computing (Neural Networks, Genetic Algorithms, etc.)
-
Visual
Exploratory Data Analysis
-
In real-time
(in the form of steering)
|
 |
" |
 |
|
Place neurons on
Geodesic dome instead of a 2-D grid - no discontinuities; any point on the
surface of the sphere will be represented by a neuron on the Geodosic Dome
wiht an equal probability. |
Visualisation of a Self-Organisation Map (SOM) in 3-D
|
 |
 |
|
Advantages: Real distance values can be used - different maps can be
compared; no cluster boundaries are obscured by skewed distribution of
distance values; the distance (in the feature space) can be measured as
geodesic distances. |
|
