/media/naith/HAlpha/python/generators/venv/bin/python3 /media/naith/HAlpha/02-DP_Self-learning_autonomous_systems_using_deferred_learning/Python/architecture_test_ff_nn.py

=== TEST: PARABOLOID (Activation: relu) ===

=== COMPARISON OF ACTIVATION FUNCTIONS ON PARABOLOID PROBLEM ===

--- Training network with activation function: relu ---

Results for activation function relu:
Training completed in 0.11 seconds (total), 247 epochs
MAE before correction (in region): 0.129822
MAE after correction (in region): 0.003160
MAX deviation before correction (in region): 0.400000
MAX deviation after correction (in region): 0.011736
Error reduction in region: 97.57%
Displaying surface: Corrected Surface - relu activation

--- Training network with activation function: sigmoid ---

Results for activation function sigmoid:
Training completed in 0.02 seconds (total), 28 epochs
MAE before correction (in region): 0.129822
MAE after correction (in region): 0.083885
MAX deviation before correction (in region): 0.400000
MAX deviation after correction (in region): 0.289919
Error reduction in region: 35.38%
Displaying surface: Corrected Surface - sigmoid activation

--- Training network with activation function: tanh ---

Results for activation function tanh:
Training completed in 0.12 seconds (total), 252 epochs
MAE before correction (in region): 0.129822
MAE after correction (in region): 0.010266
MAX deviation before correction (in region): 0.400000
MAX deviation after correction (in region): 0.043414
Error reduction in region: 92.09%
Displaying surface: Corrected Surface - tanh activation
Displaying graph: Activation Function Comparison - PARABOLOID

=== SUMMARY OF ACTIVATION FUNCTION RESULTS ===
Activation | Epochs  | Time (s)  | MAE        | MAE region   | MAX error  | Reduction (%)
-------------------------------------------------------------------------------------
relu       | 247     | 0.11      | 0.001111   | 0.003160     | 0.042160   | 97.57
sigmoid    | 28      | 0.02      | 0.008026   | 0.083885     | 0.289919   | 35.38
tanh       | 252     | 0.12      | 0.001719   | 0.010266     | 0.043414   | 92.09
Early stopping - no improvement after 20 epochs
Training completed in 0.05 seconds
Best MAE achieved: 0.002508
Final relative error: 1.93%
MAE before correction (entire surface): 0.011961
MAE after correction (entire surface): 0.001055
MAE before correction (in region): 0.129822
MAE after correction (in region): 0.002508
MAX deviation before correction (entire surface): 0.400000
MAX deviation after correction (entire surface): 0.042160
MAX deviation before correction (in region): 0.400000
MAX deviation after correction (in region): 0.009682
Error reduction in region: 98.07%
Displaying surface: 1) Stavový prostor bez odchylky (original)
Displaying surface: 2) Rozdíl mezi odezvou neuronové sítě a stavovým prostorem bez odchylky
Displaying surface: 3) Stavový prostor s odchylkou (paraboloid)
Displaying surface: 4) Chybový stavový prostor (rozdíl mezi povrchem s chybou a originálem)
Displaying surface: 5) Korigovaný povrch pomocí malé neuronové sítě (aktivace: relu)
Displaying surface: 5) Rozdíl mezi původním a opraveným povrchem (aktivace: relu)

Statistika rozdílu opraveného a původního povrchu:
Průměrná absolutní odchylka: 0.001055
Směrodatná odchylka: 0.003960
Maximální absolutní odchylka: 0.042160

Process finished with exit code 0
