深度学习模型训练结果：Epoch 1-20 训练集准确率提高至73.47%，但验证集表现欠佳

该模型经过20个时期（epoch）的训练，训练集的损失（loss）降低到0.9676，准确率（accuracy）提高到0.7347；验证集的损失为7.0724，准确率为0.0978。可以看出，该模型在训练集上表现较好，但在验证集上表现较差，存在过拟合的情况。

训练结果展示:

Epoch 1/20 457/457 [==============================] - 1549s 3s/step - loss: 3.8737 - accuracy: 0.1113 - val_loss: 3.5868 - val_accuracy: 0.1167 Epoch 2/20 457/457 [==============================] - 1380s 3s/step - loss: 3.5462 - accuracy: 0.1212 - val_loss: 3.5424 - val_accuracy: 0.1145 Epoch 3/20 457/457 [==============================] - 1484s 3s/step - loss: 3.4678 - accuracy: 0.1236 - val_loss: 3.4777 - val_accuracy: 0.1216 Epoch 4/20 457/457 [==============================] - 1487s 3s/step - loss: 3.4075 - accuracy: 0.1302 - val_loss: 3.4510 - val_accuracy: 0.1214 Epoch 5/20 457/457 [==============================] - 1501s 3s/step - loss: 3.3368 - accuracy: 0.1405 - val_loss: 3.4801 - val_accuracy: 0.1145 Epoch 6/20 457/457 [==============================] - 1371s 3s/step - loss: 3.2418 - accuracy: 0.1529 - val_loss: 3.5124 - val_accuracy: 0.1186 Epoch 7/20 457/457 [==============================] - 1408s 3s/step - loss: 3.1038 - accuracy: 0.1784 - val_loss: 3.5292 - val_accuracy: 0.1156 Epoch 8/20 457/457 [==============================] - 1440s 3s/step - loss: 2.9115 - accuracy: 0.2195 - val_loss: 3.5979 - val_accuracy: 0.1159 Epoch 9/20 457/457 [==============================] - 1399s 3s/step - loss: 2.7149 - accuracy: 0.2679 - val_loss: 3.9500 - val_accuracy: 0.1066 Epoch 10/20 457/457 [==============================] - 1563s 3s/step - loss: 2.4673 - accuracy: 0.3304 - val_loss: 3.8963 - val_accuracy: 0.1060 Epoch 11/20 457/457 [==============================] - 1593s 3s/step - loss: 2.2450 - accuracy: 0.3818 - val_loss: 4.2670 - val_accuracy: 0.1025 Epoch 12/20 457/457 [==============================] - 1584s 3s/step - loss: 2.0348 - accuracy: 0.4412 - val_loss: 4.4389 - val_accuracy: 0.1030 Epoch 13/20 457/457 [==============================] - 1565s 3s/step - loss: 1.8246 - accuracy: 0.4968 - val_loss: 4.8128 - val_accuracy: 0.1060 Epoch 14/20 457/457 [==============================] - 1570s 3s/step - loss: 1.6352 - accuracy: 0.5469 - val_loss: 4.7601 - val_accuracy: 0.0932 Epoch 15/20 457/457 [==============================] - 1596s 3s/step - loss: 1.4863 - accuracy: 0.5868 - val_loss: 5.6534 - val_accuracy: 0.0962 Epoch 16/20 457/457 [==============================] - 1660s 4s/step - loss: 1.3464 - accuracy: 0.6273 - val_loss: 5.8366 - val_accuracy: 0.0932 Epoch 17/20 457/457 [==============================] - 1487s 3s/step - loss: 1.2426 - accuracy: 0.6588 - val_loss: 5.7036 - val_accuracy: 0.0934 Epoch 18/20 457/457 [==============================] - 1486s 3s/step - loss: 1.1197 - accuracy: 0.6849 - val_loss: 6.6610 - val_accuracy: 0.0921 Epoch 19/20 457/457 [==============================] - 1518s 3s/step - loss: 1.0298 - accuracy: 0.7125 - val_loss: 6.6695 - val_accuracy: 0.0978 Epoch 20/20 457/457 [==============================] - 1614s 4s/step - loss: 0.9676 - accuracy: 0.7347 - val_loss: 7.0724 - val_accuracy: 0.0978