/Perceptron-Model-in-Different-Programming-Languages

Perceptron Model in Different Programming Languages

Primary LanguagePython

Perceptron Model in Different Programming Languages

Perceptron Model in Different Programming Languages

Description

This repository is mainly intended for educational/recreational purposes by providing example-code of a perceptron model implemented in multiple programming languages.

The example being considered is an AND gate emulation. This can be switched out to be any simple binary logic gate that's linearly separable (i.e. not XOR or NXOR) by changing the labels that are hard-coded to the desired gate.

Currently available languages:

  • Python(Numpy)
  • Python(Cython)
  • Python(Keras)
  • Python(Tensorflow)
  • Java
  • C
  • C++
  • Typescript
  • Javascript(NodeJS)
  • Go
  • Lua
  • Rust
  • Crystal
  • Swift

Future additions:

  • Javascript(p5JS)
  • C#
  • C++/CUDA
  • LuaC++

Outputs

Python(Numpy)

Weights:

[[-0.06385825]
 [ 0.31008403]]

Biases:

[[0.75214673]]


Pre-training

[[0.67964628]
 [0.74311662]
 [0.66558608]
 [0.73073848]]

Training for 100000 epochs...
Progress: 0.0% ... Error: 1.529283817045281
Progress: 10.0% ... Error: 0.5963336951291089
Progress: 20.0% ... Error: 0.3977871484781369
Progress: 30.0% ... Error: 0.3009347811294628
Progress: 40.0% ... Error: 0.24119652933532412
Progress: 50.0% ... Error: 0.20002500397421824
Progress: 60.0% ... Error: 0.1698674071810621
Progress: 70.0% ... Error: 0.14688948909543426
Progress: 80.0% ... Error: 0.12887247254702422
Progress: 90.0% ... Error: 0.11442396270617813

Post-training

[[0.01083864]
 [0.20154302]
 [0.14796895]
 [0.80002344]]

Error: 0.10262249959544203

Python(Cython)

PRE-TRAINING:
0.5000007373746488
0.5000028872163979
0.500000742147677
0.5000028919894262
Training initiated...
Epochs(): 0.000000 ... Error: 0.998758
Epochs(): 10.000000 ... Error: 0.000087
Epochs(): 20.000000 ... Error: 0.000043
Epochs(): 30.000000 ... Error: 0.000029
Epochs(): 40.000000 ... Error: 0.000021
Epochs(): 50.000000 ... Error: 0.000017
Epochs(): 60.000000 ... Error: 0.000014
Epochs(): 70.000000 ... Error: 0.000012
Epochs(): 80.000000 ... Error: 0.000011
Epochs(): 90.000000 ... Error: 0.000010
POST-TRAINING:
4.749294846102901e-09
0.0015864800965162294
0.0015864801462667664
0.9981225661409634

Python(Keras)

Epoch 1/100000
4/4 [==============================] - 0s 67ms/sample - loss: 0.2694 - acc: 0.5000
Epoch 2/100000
4/4 [==============================] - 0s 1ms/sample - loss: 0.2691 - acc: 0.5000
Epoch 3/100000
4/4 [==============================] - 0s 713us/sample - loss:0.2687 - acc: 0.5000
Epoch 4/100000
4/4 [==============================] - 0s 970us/sample - loss:0.2684 - acc: 0.5000
Epoch 5/100000
4/4 [==============================] - 0s 640us/sample - loss:0.2680 - acc: 0.5000
Epoch 6/100000
4/4 [==============================] - 0s 643us/sample - loss:0.2677 - acc: 0.5000
Epoch 7/100000
4/4 [==============================] - 0s 756us/sample - loss:0.2674 - acc: 0.5000
Epoch 8/100000
4/4 [==============================] - 0s 739us/sample - loss:0.2670 - acc: 0.5000
Epoch 9/100000
4/4 [==============================] - 0s 710us/sample - loss:0.2667 - acc: 0.5000
Epoch 10/100000
4/4 [==============================] - 0s 625us/sample - loss:0.2663 - acc: 0.5000

...

Epoch 99995/100000
4/4 [==============================] - 0s 265us/sample - loss:0.0057 - acc: 1.0000
Epoch 99996/100000
4/4 [==============================] - 0s 285us/sample - loss:0.0057 - acc: 1.0000
Epoch 99997/100000
4/4 [==============================] - 0s 339us/sample - loss:0.0057 - acc: 1.0000
Epoch 99998/100000
4/4 [==============================] - 0s 302us/sample - loss:0.0057 - acc: 1.0000
Epoch 99999/100000
4/4 [==============================] - 0s 448us/sample - loss:0.0057 - acc: 1.0000
Epoch 100000/100000
4/4 [==============================] - 0s 348us/sample - loss:0.0057 - acc: 1.0000
[[8.4083923e-04]
 [8.1297390e-02]
 [8.1297427e-02]
 [9.0296149e-01]]

Python(Tensorflow)

Epoch(%):  0.0 %    loss:  0.33210212
Epoch(%):  10.0 %    loss:  0.015254875
Epoch(%):  20.0 %    loss:  0.0068163867
Epoch(%):  30.0 %    loss:  0.0042989245
Epoch(%):  40.0 %    loss:  0.003116064
Epoch(%):  50.0 %    loss:  0.002434728
Epoch(%):  60.0 %    loss:  0.0019938578
Epoch(%):  70.0 %    loss:  0.001685899
Epoch(%):  80.0 %    loss:  0.0014588995
Epoch(%):  90.0 %    loss:  0.0012851977
Epoch(%):  100.0 %    loss:  0.0011477593

Elapsed time  30.536971328793644
________________________________________________________________________________

  [0.03386783]
  [0.03388705]
  [0.03387891]
  [0.9661194]

Java

Pre-Training
0.5348137280104677
0.723578338386435
0.7502927912229587
0.8724698405926142

Starting training for 10000000 epochs...
Epoch(%): 0%, Error: 1.3884018373317462
Epoch(%): 10%, Error: 0.010522028036657579
Epoch(%): 20%, Error: 0.0049630849565218665
Epoch(%): 30%, Error: 0.003218925932545704
Epoch(%): 40%, Error: 0.0023740418311171464
Epoch(%): 50%, Error: 0.0018773794404555915
Epoch(%): 60%, Error: 0.0015510882572958048
Epoch(%): 70%, Error: 0.0013206127992833052
Epoch(%): 80%, Error: 0.0011492857031122742
Epoch(%): 90%, Error: 0.001016999191696389
Epoch(%): 99%, Error: 9.118158525567842E-4

Post-Training
5.474909704186176E-6
0.016365370942308083
0.016365386572161952
0.9806050372816417

C

Perceptron initialised

Pre-Training
0.685033
0.840640
0.761333
0.885543

Training initiated...
Epochs(%): 0.000000 ... Error: 1.768322
Epochs(%): 10.000000 ... Error: 0.010582
Epochs(%): 20.000000 ... Error: 0.004977
Epochs(%): 30.000000 ... Error: 0.003225
Epochs(%): 40.000000 ... Error: 0.002377
Epochs(%): 50.000000 ... Error: 0.001879
Epochs(%): 60.000000 ... Error: 0.001552
Epochs(%): 70.000000 ... Error: 0.001322
Epochs(%): 80.000000 ... Error: 0.001150
Epochs(%): 90.000000 ... Error: 0.001018


Post-Training
0.000005
0.016370
0.016370
0.980600

C++

Pre-Training
0.686348 
0.764495 
0.835248 
0.88264 

Training initiated...
Epochs(%): 0 ... Error: 1.76658
Epochs(%): 10 ... Error: 0.0105841
Epochs(%): 20 ... Error: 0.00497742
Epochs(%): 30 ... Error: 0.00322505
Epochs(%): 40 ... Error: 0.0023774
Epochs(%): 50 ... Error: 0.0018795
Epochs(%): 60 ... Error: 0.00155254
Epochs(%): 70 ... Error: 0.00132167
Epochs(%): 80 ... Error: 0.00115009
Epochs(%): 90 ... Error: 0.00101763


Post-Training
5.47956e-06 
0.0163699 
0.0163699 
0.9806

Typescript

0.7304081197614447
0.7710994438929843
0.8005525165435999
0.8330760587564477
Epoch(%): 0% ... Error: 0.5932394397111913
Epoch(%): 10% ... Error: 0.00026804890681212426
Epoch(%): 20% ... Error: 0.00013127665844091298
Epoch(%): 30% ... Error: 0.00008672169316476456
Epoch(%): 40% ... Error: 0.00006468917895902311
Epoch(%): 50% ... Error: 0.000051560524953053264
Epoch(%): 60% ... Error: 0.000042850362282691576
Epoch(%): 70% ... Error: 0.000036651444473264826
Epoch(%): 80% ... Error: 0.000032015586537896924
Epoch(%): 90% ... Error: 0.00002841834265688727
1.553356647888704e-7
0.0050543214573224365
0.005054319928508864
0.9940167369413687

NodeJS

Pre-Training
0.6918047896906404
0.6998179549852507
0.7742769392294069
0.7808251531875576

Epoch(%): 0% ... Error: 0.4882705640615913
Epoch(%): 10% ... Error: 0.0031066294620815537
Epoch(%): 20% ... Error: 0.0014612005571914071
Epoch(%): 30% ... Error: 0.0009469467937243809
Epoch(%): 40% ... Error: 0.000698162821943197
Epoch(%): 50% ... Error: 0.0005520079884997068
Epoch(%): 60% ... Error: 0.0004560234209277485
Epoch(%): 70% ... Error: 0.0003882400838516292
Epoch(%): 80% ... Error: 0.000337859994030239
Epoch(%): 90% ... Error: 0.00029896418624558594

Post-Training
0.000005481532567786762
0.016371928302539418
0.016371877905484358
0.9805974566723038

Go

0.6602839959986757
0.8327256756202939
0.7806092539693458
0.9011198208020211
0% ... Error: 1.748535
10% ... Error: 0.010564
20% ... Error: 0.004973
30% ... Error: 0.003223
40% ... Error: 0.002376
50% ... Error: 0.001879
60% ... Error: 0.001552
70% ... Error: 0.001321
80% ... Error: 0.001150
90% ... Error: 0.001017
100% ... Error: 0.000912
5.478050537819302e-06
0.016368443759505653
0.01636845939806802
0.9806013895847836

Lua (5.1.5)

Pre-Training
0.68634768291945
0.76449478576855
0.83524794042728
0.88263985343676
Epoch(%): 0%, Error: 0.46992565191389
Epoch(%): 10%, Error: 6.0658760192174e-08
Epoch(%): 20%, Error: 5.6186250023124e-09
Epoch(%): 30%, Error: 1.4556711800043e-09
Epoch(%): 40%, Error: 5.6686729161738e-10
Epoch(%): 50%, Error: 2.7482642541124e-10
Epoch(%): 60%, Error: 1.5277640500465e-10
Epoch(%): 70%, Error: 9.3254877141364e-11
Epoch(%): 80%, Error: 6.0925148702575e-11
Epoch(%): 90%, Error: 4.1910761337228e-11
Post-Training
5.4792396891681e-06
0.016369648481205
0.016369701503219
0.98060015144136

Rust

Pre-Training
0.7310585786300049
0.8807970779778823
0.8807970779778823
0.9525741268224334

Epoch(%): 0% ... Error: 0.522014223238681
Epoch(%): 10% ... Error: 0.0026408552351573234
Epoch(%): 20% ... Error: 0.001243168518649312
Epoch(%): 30% ... Error: 0.0008057578303075101
Epoch(%): 40% ... Error: 0.000594074621917179
Epoch(%): 50% ... Error: 0.00046970020535868696
Epoch(%): 60% ... Error: 0.00038801591902687887
Epoch(%): 70% ... Error: 0.0003303307068047589
Epoch(%): 80% ... Error: 0.0002874563173784957
Epoch(%): 90% ... Error: 0.0002543557209483691
Epoch(%): 99% ... Error: 0.00022803931282502183

Post-Training
0.0000054780388623871445
0.01636843253816961
0.016368448176693976
0.9806014036120793

Crystal (0.27.2)

[0.0, 0.0]
0.6908712064682041
[0.0, 1.0]
0.7747915745712539
[1.0, 0.0]
0.8452513342138535
[1.0, 1.0]
0.8937094231025183
0.0% ... Error: 0.45083812817023344
10.0% ... Error: 0.03822879901009097
20.0% ... Error: 0.018729354312600907
30.0% ... Error: 0.011963078000812529
40.0% ... Error: 0.008658574297653066
50.0% ... Error: 0.006733627001739041
60.0% ... Error: 0.005484893652388013
70.0% ... Error: 0.004614023985002622
80.0% ... Error: 0.003974307661524453
90.0% ... Error: 0.0034856894292992517
100.0% ... Error: 0.0031009707769695505
[0.0, 0.0]
0.0003241921893645471
[0.0, 1.0]
0.06361693593303154
[1.0, 0.0]
0.06361722141257974
[1.0, 1.0]
0.9343531427983828

Swift (4.2.1)

Pre-Training:
0.5871688281141535
0.6488127336921564
0.7849738529623651
0.8258421575372155

Epochs(%): 0 ... Error: 0.3529676724506121
Epochs(%): 10 ... Error: 0.005687036001512336
Epochs(%): 20 ... Error: 0.002640473929544105
Epochs(%): 30 ... Error: 0.001696162404866232
Epochs(%): 40 ... Error: 0.0012430738240744982
Epochs(%): 50 ... Error: 0.0009785707575491922
Epochs(%): 60 ... Error: 0.0008057156923088166
Epochs(%): 70 ... Error: 0.0006841279984251465
Epochs(%): 80 ... Error: 0.0005940508381222067
Epochs(%): 90 ... Error: 0.0005246967351030535
Epochs(%): 100 ... Error: 0.0004696849220425556

Post-Training:
1.6569391297523873e-05
0.02348413831843244
0.023484183755364978
0.9721481100388056