下面是keras訓練模型的記錄，epoch=20，不管我怎么修改模型總會在第12個epoch左右測驗集出現損失明顯高的現象，請問這是什么原因？我用的損失函式是MSE，可能對于脈沖噪聲起到了放大作用。但問題是為什么不是每個epoch的MSE都很高，而是訓練到第12個epoch的時候突然增高？
Epoch 1/20
60/60 [==============================] - 167s 3s/step - loss: 0.8877 - val_loss: 0.2337
Epoch 2/20
60/60 [==============================] - 166s 3s/step - loss: 0.7081 - val_loss: 0.1796
Epoch 3/20
60/60 [==============================] - 166s 3s/step - loss: 0.6973 - val_loss: 0.1650
Epoch 4/20
60/60 [==============================] - 168s 3s/step - loss: 0.6907 - val_loss: 0.0897
Epoch 5/20
60/60 [==============================] - 167s 3s/step - loss: 0.6847 - val_loss: 0.1025
Epoch 6/20
60/60 [==============================] - 164s 3s/step - loss: 0.6809 - val_loss: 0.1731
Epoch 7/20
60/60 [==============================] - 167s 3s/step - loss: 0.6770 - val_loss: 0.2273
Epoch 8/20
60/60 [==============================] - 167s 3s/step - loss: 0.6744 - val_loss: 0.1221
Epoch 9/20
60/60 [==============================] - 167s 3s/step - loss: 0.6724 - val_loss: 0.5085
Epoch 10/20
60/60 [==============================] - 166s 3s/step - loss: 0.6713 - val_loss: 0.4322
Epoch 11/20
60/60 [==============================] - 165s 3s/step - loss: 0.6695 - val_loss: 0.2419
Epoch 12/20
60/60 [==============================] - 167s 3s/step - loss: 0.7354 - val_loss: 11.4644
Epoch 13/20
60/60 [==============================] - 165s 3s/step - loss: 0.6667 - val_loss: 4.5552
Epoch 14/20
60/60 [==============================] - 164s 3s/step - loss: 0.6653 - val_loss: 0.0955
Epoch 15/20
60/60 [==============================] - 165s 3s/step - loss: 0.6629 - val_loss: 0.0924
Epoch 16/20
60/60 [==============================] - 165s 3s/step - loss: 0.6596 - val_loss: 0.1004
Epoch 17/20
60/60 [==============================] - 167s 3s/step - loss: 0.6558 - val_loss: 0.3345
Epoch 18/20
60/60 [==============================] - 167s 3s/step - loss: 0.7868 - val_loss: 0.3541
Epoch 19/20
60/60 [==============================] - 168s 3s/step - loss: 0.6708 - val_loss: 0.2062
Epoch 20/20
60/60 [==============================] - 166s 3s/step - loss: 0.6507 - val_loss: 0.1870
轉載請註明出處，本文鏈接：https://www.uj5u.com/qita/31511.html
標籤：人工智能技術
上一篇：神經網路的性能評判指標有哪些？
下一篇：你們說搞一博客有必要嗎？