Page 99 - Data Science Algorithms in a Week

Evolutionary Optimization of Support Vector Machines … 83

 Pointalization – Describes the size of the points that make the individual circles
(3 levels).
 Saturation – Describes the strength of the color within the circles (3 levels).
 Brightness – Describes the amount of light in the circles themselves (4 levels).
 Blur – Describes the crispness of the circles (2 levels).
 Background – Describes the background color of the image (3 levels).

Table 1. Features used to generate the 624 images (Rabelo, 2000)

Attribute Level 1 Level 2 Level 3 Level 4
1 Density X3 X2 X1 --
2 Color family Cold: blue, green purples Warm: red, orange --
3 Pointalization 5 15 50 --
4 Saturation 50 0 -- --
5 Brightness 50 -- -- -25
6 Motion blur 0 10 -- --
7 Background Black Gray White --

Density: Level 1 Cold vs Warm: Level 1 Density: Level 1 Cold vs Warm: Level 1
Pointalized: Level 1 Saturation: Level 1 Pointalized: Level 1 Saturation: Level 1
Light/Dark: Level 1 Motion blur: Level 1 Light/Dark: Level 2 Motion blur: Level 2
BKG: Level 3 BKG: Level 3
Figure 7. Images with features 1111113 and 1111223, respectively.

Density: Level 1 Cold vs Warm: Level 2 Density: Level 1 Cold vs Warm: Level 2
Pointalized: Level 1 Saturation: Level 1 Pointalized: Level 3 Saturation: Level 1
Light/Dark: Level 3 Motion blur: Level 2 Light/Dark: Level 3 Motion blur: Level 1
BKG: Level 2 BKG: Level 1
Figure 8. Images with features 1211323 and 1231311, respectively.