Neural Network Experiment 2

A visual representation of a simple neural network expanding on previous work done on https://github.com/samueleishion/neural-network-experiment

How to run it

make network

To clean the project:

make clean

How to interact

Several nodes (neurons) will be displayed. The top ones are sensorial, the middle ones are transmitters, and the bottom ones are terminals. You can click on the sensorials to transmit signals.

On the right side of the graphic window, there is a graph/map-like visualization of the interactions between sensorial and terminal neurons. The sensorial neurons are displayed on the y-axis while the terminal neurons are displayed on the x-axis.

Several features can be changed to change the experiment on the settings.py file. The features include:

Window Size

WINDOW_X = 1000

purpose: establishes width of graphic window
type: int
default value: 1000px

WINDOW_Y = 600

purpose: establishes height of graphic window
type: int
default value: 600px

Number of neurons

SENSORY_NEURONS = 10

purpose: establishes total number of sensorial neurons
type: int
default value: 10

TERMINAL_NEURONS = 10

purpose: establishes total number of terminal neurons
type: int
default value: 10

Sensation intensity

INTENSITY = 20

purpose: equivalent to duration/potence of sensation, this is a fixed number for all sensations in this experiment
type: int
default value: 20

Automatic and monitored interaction

AUTOMATIC = False

purpose: allows the sensations to be automatically distributed
type: boolean
default value: False
detail: setting this feature to True facilitates the graph visualization and experimentation

GRAPH = True

purpose: displays the right panel with the graph/map-like visualization
type: boolean
default value: True
detail: this graph maps the number of reached terminals after trigerring a given sensory neuron.

Visualization

If the GRAPH feature is set to True, the y-axis represents the sensorial neurons. The x-axis represents the sensorial neurons. If a given sensory neuron (S) has sent a transmission and is received by a given terminal neuron (T), then the graph at coordinates (S,T) will be lit up with a shade of red (given that, for this particular graphics window the y-axis increases downwards).

For example, we have 3 sensory neurons and 4 terminal neurons. This would be a numeric representation of the graph:

	T1	T2	T3	T4
S1	0	0	0	0
S2	0	0	0	0
S3	0	0	0	0

After several interactions, say that:

S1 has successfully transmitted
- 3 sensations to T1
- 2 sensations to T2
S2 has successfully transmitted
- 1 sensation to T2
- 3 sensations to T3
- 2 sensations to T4
S3 has successfully transmitted
- 2 sensations to T1
- 4 sensations to T3
- 1 sensation to T4

Our graph will now be:

	T1	T2	T3	T4
S1	.6	.4	0	0
S2	0	.17	.5	.33
S3	.29	0	.57	.14

On the actual visualization, these decimals are used to calculate the intensity of the color.

samueleishion / neural-network-experiment-2