Table 2 Test data used to evaluate the interpolation capabilities of the trained neural network.

From: Neural network based formation of cognitive maps of semantic spaces and the putative emergence of abstract concepts

Name

Height (cm)

Weight (kg)

Number legs

Danger (subjective)

Reproduction (2 = Birth, 1 = Eggs)

Fur (2 = No, 1 = Yes)

Lungs (2 = No, 1 = Yes)

Jaguar

70

70

4

90

2

1

1

Donkey

100

200

4

10

2

1

1

Wild boar

70

180

4

20

2

1

1

Melontha

2.5

0.001

6

0

1

2

2

Dragonfly

6

0.0003

6

0

1

2

2

Wasp

1.5

0.00008

6

20

1

2

2

  1. It consists of 6 different animal species, which belong to three different taxonomic classes: mammals, insects and amphibians. Again, each animal is characterized by seven semantic features: Height, weight, number of legs, its danger level, the reproduction system, if it has fur and if it has lungs.
Back to article page