ANN Simulator Sandbox

ANN Simulator Sandbox preview image

1 collaborator

Larry_bencivengo Larry Bencivengo (Author)

Tags

artificial intelligence 

Tagged by Larry Bencivengo almost 4 years ago

artificial neural network 

Tagged by Larry Bencivengo almost 4 years ago

Visible to everyone | Changeable by the author
Model was written in NetLogo 6.1.1 • Viewed 325 times • Downloaded 22 times • Run 0 times
Download the 'ANN Simulator Sandbox' modelDownload this modelEmbed this model

Do you have questions or comments about this model? Ask them here! (You'll first need to log in.)


Comments and Questions

Please start the discussion about this model! (You'll first need to log in.)

Click to Run Model

;; Artificial Neural Network Sandbox
;;
;; by L. F. Bencivengo, Jr.  December 14, 2020

globals [OutputA OutputB  OutputC
         OutputI OutputII OutputIII
         Output1 Output2  Output3
         started? auto_Input_Pattern input#]

breed [Input_Neurons Input_Neuron]
breed [Hidden_Neurons Hidden_Neuron]
breed [Output_Neurons Output_Neuron]
turtles-own [threshold weight1 weight2 weight3 activation output]

directed-link-breed [axons-AI axon-AI]
directed-link-breed [axons-BII axon-BII]
directed-link-breed [axons-CIII axon-CIII]

;; BUTTONS: This section defines the function of each Interface Button

to setup
  clear-all
  ;; turn off auto_Input_Pattern
  set auto_Input_Pattern FALSE
  set input# 0
  ;; create neurons with current output = 0
  set-default-shape Input_Neurons "square 2"
  set-default-shape Hidden_Neurons "triangle 2"
  set-default-shape Output_Neurons "circle 2"
  setup-Input_Neurons
  setup-Hidden_Neurons
  setup-Output_Neurons
  ;; create axons that connect each neuron in the network
  set-default-shape axons-AI "axon-ai-0"
  set-default-shape axons-BII "axon-bii-0"
  set-default-shape axons-CIII "axon-ciii-0"
  setup-synapses
  ;; clear the ticks to disable Input Buttons
  set started? FALSE
  clear-ticks
end 

to go
  ;; if just starting, reset ticks and enable Input Buttons
  if not started? [reset-ticks reset-timer]
  set started? TRUE
  ;; read values of weights and thresholds from sliders
  update-weights
  ;; update the activation and output of each neuron in the network
  update-Hidden_Outputs
  update-Output_Outputs
  ;; update the display to show any changes in neurons' output
  update-display
  tick
end 

to auto-Input_Pattern
  ;; if auto-input is clicked, run through the auto-input sequence
  ;; through repeated calls of update-Input_Pattern
  ;; Forever Button, Disabled until ticks start
end 

to Input_A
  ;; change the current state of Input Neuron A
  ifelse OutputA = 0 [set OutputA 1] [set OutputA 0]
  ifelse OutputA = 0 [ask turtle 0 [set output 0]] [ask turtle 0 [set output 1]]
  stop
end 

to Input_B
  ;; change the current state of Input Neuron B
  ifelse OutputB = 0 [set OutputB 1] [set OutputB 0]
  ifelse OutputB = 0 [ask turtle 1 [set output 0]] [ask turtle 1 [set output 1]]
end 

to Input_C
  ;; change the current state of Input Neuron C
  ifelse OutputC = 0 [set OutputC 1] [set OutputC 0]
  ifelse OutputC = 0 [ask turtle 2 [set output 0]] [ask turtle 2 [set output 1]]
end 

to reset_weights
  ;; reset the sliders for every neuron's weights and thresholds to the default value of 1
  set weightIA 1 set weightIB 1 set weightIC 1
  set thresholdI 1
  set weightIIA 1 set weightIIB 1 set weightIIC 1
  set thresholdII 1
  set weightIIIA 1 set weightIIIB 1 set weightIIIC 1
  set thresholdIII 1
  set weight1I 1 set weight1II 1 set weight1III 1
  set threshold1 1
  set weight2I 1 set weight2II 1 set weight2III 1
  set threshold2 1
  set weight3I 1 set weight3II 1 set weight3III 1
  set threshold3 1
end 

;; PROCEDURES: This section defines the Procedures which control the model

to setup-Input_Neurons
  ;; turtle 0 is Input Neuron A
  create-Input_Neurons 1 [
    set shape "input neuron a - 0"
    set size 3  set color grey
    set xcor -8  set ycor 12]
  ;; turtle 1 is Input Neuron B
  create-Input_Neurons 1 [
    set shape "input neuron b - 0"
    set size 3  set color grey
    set xcor 0  set ycor 12]
  ;; turtle 2 is Input Neuron C
  create-Input_Neurons 1 [
    set shape "input neuron c - 0"
    set size 3  set color grey
    set xcor 8  set ycor 12]
end 

to setup-Hidden_Neurons
  ;; turtle 3 is Hidden Neuron I
  create-Hidden_Neurons 1 [
    set shape "hidden neuron i - 0"
    set size 3  set color grey
    set xcor -8  set ycor 0]
  ;; turtle 4 is Hidden Neuron II
  create-Hidden_Neurons 1 [
    set shape "hidden neuron ii - 0"
    set size 3  set color grey
    set xcor 0  set ycor 0]
  ;; turtle 5 is Hidden Neuron III
  create-Hidden_Neurons 1 [
    set shape "hidden neuron iii - 0"
    set size 3  set color grey
    set xcor 8  set ycor 0]
end 

to setup-Output_Neurons
  ;; turtle 6 is Output Neuron 1
  create-Output_Neurons 1 [
    set shape "output neuron 1 - 0"
    set size 3  set color grey
    set xcor -8  set ycor -12]
  ;; turtle 7 is Output Neuron 2
  create-Output_Neurons 1 [
    set shape "output neuron 2 - 0"
    set size 3  set color grey
    set xcor 0  set ycor -12]
  ;; turtle 8 is Output Neuron 3
  create-Output_Neurons 1 [
    set shape "output neuron 3 - 0"
    set size 3  set color grey
    set xcor 8  set ycor -12]
end 

to setup-synapses
  ;; create axons from each Input Neuron to each Hidden Neuron
  ask turtle 0 [create-axons-AI-to Hidden_Neurons]
  ask turtle 1 [create-axons-BII-to Hidden_Neurons]
  ask turtle 2 [create-axons-CIII-to Hidden_Neurons];; [set thickness 0.25 set color white]]
  ;; create axons from each Hidden Neuron to each Output Neuron
  ask turtle 3 [create-axons-AI-to Output_Neurons]
  ask turtle 4 [create-axons-BII-to Output_Neurons]
  ask turtle 5 [create-axons-CIII-to Output_Neurons]
end 

to update-Input_Pattern
  ;; when auto_Input_Pattern = TRUE, runs through a sequence of input patterns
  ;; changing the pattern at a rate that depends upon the auto-Input_Rate slider value
    ifelse input# = 8 [set input# 1] [set input# input# + 1]
    (ifelse
      input# = 1 [ask turtle 0 [set output 0] ask turtle 1 [set output 0] ask turtle 2 [set output 0]
                  set OutputA 0 set OutputB 0 set OutputC 0]
      input# = 2 [ask turtle 0 [set output 1] ask turtle 1 [set output 0] ask turtle 2 [set output 0]
                  set OutputA 1 set OutputB 0 set OutputC 0]
      input# = 3 [ask turtle 0 [set output 0] ask turtle 1 [set output 1] ask turtle 2 [set output 0]
                  set OutputA 0 set OutputB 1 set OutputC 0]
      input# = 4 [ask turtle 0 [set output 0] ask turtle 1 [set output 0] ask turtle 2 [set output 1]
                  set OutputA 0 set OutputB 0 set OutputC 1]
      input# = 5 [ask turtle 0 [set output 1] ask turtle 1 [set output 1] ask turtle 2 [set output 0]
                  set OutputA 1 set OutputB 1 set OutputC 0]
      input# = 6 [ask turtle 0 [set output 1] ask turtle 1 [set output 0] ask turtle 2 [set output 1]
                  set OutputA 1 set OutputB 0 set OutputC 1]
      input# = 7 [ask turtle 0 [set output 0] ask turtle 1 [set output 1] ask turtle 2 [set output 1]
                  set OutputA 0 set OutputB 1 set OutputC 1]
      input# = 8 [ask turtle 0 [set output 1] ask turtle 1 [set output 1] ask turtle 2 [set output 1]
                  set OutputA 1 set OutputB 1 set OutputC 1]
                 [])
  reset-timer
end 

to update-weights
  ;; read weights from slider values:
  ;; Ex. "weight1" is a neuron variable while "weightIA" is the value read from a slider
  ;; update Hidden Neuron weights
  ask turtle 3 [set weight1 weightIA set weight2 weightIB set weight3 weightIC set threshold thresholdI]
  ask turtle 4 [set weight1 weightIIA set weight2 weightIIB set weight3 weightIIC set threshold thresholdII]
  ask turtle 5 [set weight1 weightIIIA set weight2 weightIIIB set weight3 weightIIIC set threshold thresholdIII]
  ;; update Output Neuron weights
  ask turtle 6 [set weight1 weight1I set weight2 weight1II set weight3 weight1III set threshold threshold1]
  ask turtle 7 [set weight1 weight2I set weight2 weight2II set weight3 weight2III set threshold threshold2]
  ask turtle 8 [set weight1 weight3I set weight2 weight3II set weight3 weight3III set threshold threshold3]
end 

to update-Hidden_Outputs
  ;; multiply the value of each input to a Hidden Neuron by the weight of its synapse
  ;; compare the weighted sum to the threshold to determine the neuron's output (0 or 1)
  ask turtle 3 [
    set activation (weight1 * OutputA + weight2 * OutputB + weight3 * OutputC)
    ifelse activation > threshold [set output 1] [set output 0]
    ifelse output = 1 [set OutputI 1] [set OutputI 0]
  ]
  ask turtle 4 [
    set activation (weight1 * OutputA + weight2 * OutputB + weight3 * OutputC)
    ifelse activation > threshold [set output 1] [set output 0]
    ifelse output = 1 [set OutputII 1] [set OutputII 0]
  ]
  ask turtle 5 [
    set activation (weight1 * OutputA + weight2 * OutputB + weight3 * OutputC)
    ifelse activation > threshold [set output 1] [set output 0]
    ifelse output = 1 [set OutputIII 1] [set OutputIII 0]
  ]
end 

to update-Output_Outputs
  ;; multiply the value of each input to an Ouput Neuron by the weight of its synapse
  ;; compare the weighted sum to the threshold to determine the neuron's output (0 or 1)
  ask turtle 6 [
    set activation (weight1 * OutputI + weight2 * OutputII + weight3 * OutputIII)
    ifelse activation > threshold [set output 1] [set output 0]
    ifelse output = 1 [set Output1 1] [set Output1 0]
  ]
  ask turtle 7 [
    set activation (weight1 * OutputI + weight2 * OutputII + weight3 * OutputIII)
    ifelse activation > threshold [set output 1] [set output 0]
    ifelse output = 1 [set Output2 1] [set Output2 0]
  ]
  ask turtle 8 [
    set activation (weight1 * OutputI + weight2 * OutputII + weight3 * OutputIII)
    ifelse activation > threshold [set output 1] [set output 0]
    ifelse output = 1 [set Output2 1] [set Output2 0]
  ]
end 

to update-display
  ;; changes each neuron's shape and color to display whether its output is 0 or 1
  ask turtle 0 [
    ifelse output = 1
    [set shape "input neuron a - 1" set color white ask my-out-links [set color white]]
    [set shape "input neuron a - 0" set color grey ask my-out-links [set color grey]]
  ]
  ask turtle 1 [
    ifelse output = 1
    [set shape "input neuron b - 1" set color white ask my-out-links [set color white]]
    [set shape "input neuron b - 0" set color grey ask my-out-links [set color grey]]
  ]
  ask turtle 2 [
    ifelse output = 1
    [set shape "input neuron c - 1" set color white ask my-out-links [set color white]]
    [set shape "input neuron c - 0" set color grey ask my-out-links [set color grey]]
  ]
  ask turtle 3 [
    ifelse output = 1
    [set shape "hidden neuron i - 1" set color white ask my-out-links [set color white]]
    [set shape "hidden neuron i - 0" set color grey ask my-out-links [set color grey]]
  ]
  ask turtle 4 [
    ifelse output = 1
    [set shape "hidden neuron ii - 1" set color white ask my-out-links [set color white]]
    [set shape "hidden neuron ii - 0" set color grey ask my-out-links [set color grey]]
  ]
  ask turtle 5 [
    ifelse output = 1
    [set shape "hidden neuron iii - 1" set color white ask my-out-links [set color white]]
    [set shape "hidden neuron iii - 0" set color grey ask my-out-links [set color grey]]
  ]
  ask turtle 6 [
    ifelse output = 1
    [set shape "output neuron 1 - 1" set color white ask my-out-links [set color white]]
    [set shape "output neuron 1 - 0" set color grey ask my-out-links [set color grey]]
  ]
  ask turtle 7 [
    ifelse output = 1
    [set shape "output neuron 2 - 1" set color white ask my-out-links [set color white]]
    [set shape "output neuron 2 - 0" set color grey ask my-out-links [set color grey]]
  ]
  ask turtle 8 [
    ifelse output = 1
    [set shape "output neuron 3 - 1" set color white ask my-out-links [set color white]]
    [set shape "output neuron 3 - 0" set color grey ask my-out-links [set color grey]]
  ]
end 

There are 3 versions of this model.

Uploaded by When Description Download
Larry Bencivengo almost 4 years ago added labels to neurons Download this version
Larry Bencivengo almost 4 years ago added labels to neurons Download this version
Larry Bencivengo almost 4 years ago Initial upload Download this version

Attached files

File Type Description Last updated
ANN Simulator Sandbox.png preview preview image almost 4 years ago, by Larry Bencivengo Download
ANN Simulator Sandbox.png preview Preview for 'ANN Simulator Sandbox' almost 4 years ago, by Larry Bencivengo Download

This model does not have any ancestors.

This model does not have any descendants.