/simple-RNN-on-Trigonometric-circle

Primary LanguagePython

simple-RNN-on-Trigonometric-circle

My model tries to learn Trigonometric functions.

let's check the results...

Sine function :

last five epochs :

Epoch 95/100
75/75 [==============================] - 0s 3ms/step - loss: 0.0024
Epoch 96/100
75/75 [==============================] - 0s 3ms/step - loss: 0.0018
Epoch 97/100
75/75 [==============================] - 0s 3ms/step - loss: 0.0013
Epoch 98/100
75/75 [==============================] - 0s 3ms/step - loss: 0.0017
Epoch 99/100
75/75 [==============================] - 0s 3ms/step - loss: 0.0021
Epoch 100/100
75/75 [==============================] - 0s 3ms/step - loss: 0.0020
16/16 [==============================] - 0s 2ms/step - loss: 0.0010

model.evaluate : 0.0010310098296031356

Sine

Cosine function :

last five epochs :

Epoch 95/100
75/75 [==============================] - 0s 3ms/step - loss: 0.0019
Epoch 96/100
75/75 [==============================] - 0s 3ms/step - loss: 0.0020
Epoch 97/100
75/75 [==============================] - 0s 3ms/step - loss: 0.0017
Epoch 98/100
75/75 [==============================] - 0s 3ms/step - loss: 0.0021
Epoch 99/100
75/75 [==============================] - 0s 3ms/step - loss: 0.0021
Epoch 100/100
75/75 [==============================] - 0s 3ms/step - loss: 0.0021
16/16 [==============================] - 0s 2ms/step - loss: 0.0020

model.evaluate : 0.0020051589235663414

Cosine