Page 192 - AI Computer 10

AI Activity
AI Activity
Human Neural Network
Human Neural Network
In the chapter, you have learnt about the basics of neural networks. Now, let us learn more about
neural networks through a game based on creating a human neural network.
With the help of the Human Neural Network, students can easily grasp the working procedure of a
neural network. To create a human neural network, 20 students are required. These students will
act like a node of Input Layer, Hidden Layer 1, Hidden Layer 2, and Output Layer.
These 20 students are arranged in such a way that 7 students are standing in one row considered as
Input Layer, 6 students in next row considered as Hidden Layer 1, 6 students in next row considered
as Hidden Layer 2, and 1 student in the last row considered as the Output Layer.
Layers Number of student Number of chits

Input 7 6
Hidden Layer 1 6 4
Hidden Layer 2 6 2
Output Layer 1 -
TOTAL 20 -

The diagrammatic representation of these students are shown in the figure given below:

Input Hidden Hidden Output
Layer Layer 1 Layer 2 Layer

Some basic rules must be followed during the game. These rules are:
a. No communication between students, whether of the same row or different row, is allowed.
b. An appropriate distance is needed between each row.
c. Chits should be assigned to each student on which they can write different words.

d. No sentences or multiple words are to be written on the chit. Only one word per chit is allowed.
Game Instructions:

a. An image should be shown to each student of the input layer. Each student must write 6 different
words related to the shown image, one on each chit.

b. After writing their chits, each student of the input layer needs to pass on one chit to each of the
nodes of the Hidden Layer 1. In this way, each student of Hidden Layer 1 gets 7 chits from 7
different input nodes.

c. Each student of Hidden Layer 1 needs to study their chits and write down 4 different words, one
on each chit of their own. They can either use the same words as the input layer or they can make
their own description.

58
58