Neural Network
Kecerdasan Buatan
Ferry Ariyanto
1122800020
I1
I2
H1
H2
C
1 1
W1 0.1
W2 0.2
W3 0.3
W4 0.4
W5 0.5
W6 0.6
W7 0.7
W8 0.8
W9 0.9
Learning Rate = 0.2
Epoch 1 Data 1
Forward
Net H1 0.5
Net H2 0.2
Out H1 0.622459331
Out H2 0.549833997
Net C 1.692818063
Out C 0.844594403
Squared Error 0.71334
New Weight
W1 0.095903619
W2 0.19512789
W3 0.3
W4 0.4
W5 0.5
W6 0.6
W7 0.677828603
W8 0.786199207
W9 0.887809412
Epoch 5 Data 1
Forward
Net H1 0.489880679
Net H2 0.17413205
Out H1 0.62008323
Out H2 0.543423345
Net C 1.433796638
Out C 0.807492187
Squared Error 0.65204
New Weight
W1 0.072358788
W2 0.169185798
W3 0.29035259
W4 0.390929085
W5 0.48988068
W6 0.590404503
W7 0.532198544
W8 0.690074757
W9 0.794089312
Epoch 9 Data 1
Forward
Net H1 0.48025404
Net H2 0.14988401
Out H1 0.617807861
Out H2 0.53740101
Net C 1.433796638
Out C 0.807492187
Squared Error 0.65204
New Weight
W1 0.050872608
W2 0.145099991
W3 0.28129087
W4 0.382331105
W5 0.48025404
W6 0.581188886
W7 0.417462434
W8 0.614764176
W9 0.720745821
Epoch 13 Data 1
Forward
Net H1 0.471986873
Net H2 0.12922536
Out H1 0.615853915
Out H2 0.532261457
Net C 0.900947915
Out C 0.711144261
Squared Error 0.50573
New Weight
W1 0.033211725
W2 0.124851542
W3 0.27364825
W4 0.375042047
W5 0.47198687
W6 0.57323053
W7 0.229322268
W8 0.491985951
W9 0.601320363
Epoch 15 Data 1
Forward
Net H1 0.46856638
Net H2 0.12070407
Out H1 0.615044382
Out H2 0.530139432
Net C 0.779320643
Out C 0.685533678
Squared Error 0.46996
New Weight
W1 0.026162448
W2 0.116599631
W3 0.27054852
W4 0.372093604
W5 0.46856638
W6 0.569944296
W7 0.15988765
W8 0.446868301
W9 0.557484016
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
0 2 4 6 8 10 12 14 16
Gradient Descent
SSE MSE

NN Backprop Calculation.pdf

  • 1.
  • 2.
    I1 I2 H1 H2 C 1 1 W1 0.1 W20.2 W3 0.3 W4 0.4 W5 0.5 W6 0.6 W7 0.7 W8 0.8 W9 0.9 Learning Rate = 0.2
  • 3.
    Epoch 1 Data1 Forward Net H1 0.5 Net H2 0.2 Out H1 0.622459331 Out H2 0.549833997 Net C 1.692818063 Out C 0.844594403 Squared Error 0.71334 New Weight W1 0.095903619 W2 0.19512789 W3 0.3 W4 0.4 W5 0.5 W6 0.6 W7 0.677828603 W8 0.786199207 W9 0.887809412
  • 4.
    Epoch 5 Data1 Forward Net H1 0.489880679 Net H2 0.17413205 Out H1 0.62008323 Out H2 0.543423345 Net C 1.433796638 Out C 0.807492187 Squared Error 0.65204 New Weight W1 0.072358788 W2 0.169185798 W3 0.29035259 W4 0.390929085 W5 0.48988068 W6 0.590404503 W7 0.532198544 W8 0.690074757 W9 0.794089312
  • 5.
    Epoch 9 Data1 Forward Net H1 0.48025404 Net H2 0.14988401 Out H1 0.617807861 Out H2 0.53740101 Net C 1.433796638 Out C 0.807492187 Squared Error 0.65204 New Weight W1 0.050872608 W2 0.145099991 W3 0.28129087 W4 0.382331105 W5 0.48025404 W6 0.581188886 W7 0.417462434 W8 0.614764176 W9 0.720745821
  • 6.
    Epoch 13 Data1 Forward Net H1 0.471986873 Net H2 0.12922536 Out H1 0.615853915 Out H2 0.532261457 Net C 0.900947915 Out C 0.711144261 Squared Error 0.50573 New Weight W1 0.033211725 W2 0.124851542 W3 0.27364825 W4 0.375042047 W5 0.47198687 W6 0.57323053 W7 0.229322268 W8 0.491985951 W9 0.601320363
  • 7.
    Epoch 15 Data1 Forward Net H1 0.46856638 Net H2 0.12070407 Out H1 0.615044382 Out H2 0.530139432 Net C 0.779320643 Out C 0.685533678 Squared Error 0.46996 New Weight W1 0.026162448 W2 0.116599631 W3 0.27054852 W4 0.372093604 W5 0.46856638 W6 0.569944296 W7 0.15988765 W8 0.446868301 W9 0.557484016
  • 8.
    0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 0 2 46 8 10 12 14 16 Gradient Descent SSE MSE