简体繁体 English

为什么验证准确度（或训练准确度）与数据集的数量不完全匹配？

[英]why the validation accuracy (or training accuracy) is not exactly matched with the number of dataset?

原文 2020-06-29 05:19:05 0 1 python/ tensorflow/ keras/ cnn/ resnet

I'm currently using Keras to training ResNet model with dog and cat dataset.我目前正在使用 Keras 用狗和猫数据集训练 ResNet model。 For example, if I have 4000 training sets, then I think the training accuracy should be 1/4000 or 0 or the multiples of 1/4000.例如，如果我有 4000 个训练集，那么我认为训练准确率应该是 1/4000 或 0 或 1/4000 的倍数。 but the accuracy shown in the prompt is not multiples of 1/4000 and 0. how does this happen?但是提示中显示的精度不是1/4000和0的倍数，这是怎么回事呢？

enter image description here在此处输入图像描述

1 个解决方案

I don't know for sure, but I think the problem is related to the fact your lass is nan and the accuracy is 1. I will recommend you to check your hyperparameters like the "lass function" or the "batch size".我不确定，但我认为问题与您的 lass 为 nan 且准确度为 1 的事实有关。我建议您检查您的超参数，如“lass 函数”或“批量大小”。

In any case, good luck!无论如何，祝你好运！

为什么训练准确率和验证准确率会波动并缓慢增加？ - why training accuracy and validation accuracy oscillate and increase very slowly?

为什么在第一个 epoch 验证准确度高于训练准确度？ - Why at first epoch validation accuracy is higher than training accuracy?

当数据集用于验证和训练时，为什么它的准确性很差？ - Why there's a bad accuracy on dataset when it's used both for validation and training?

训练准确度好但验证准确度差 - Good training accuracy but poor validation accuracy

验证精度远低于训练精度 - Validation accuracy is much less than Training accuracy

训练精度低于验证精度 - Training accuracy is less than validation accuracy

良好的训练/验证准确度，但测试准确度差 - Good training/validation accuracy but poor test accuracy

验证准确率非常低，但训练准确率很高 - Very low validation accuracy but high training accuracy

为什么训练准确率会波动？ - Why is the training accuracy fluctuating?

为什么训练时准确性和损失保持不变？ - Why is accuracy and loss staying exactly the same while training?

暂无

暂无

声明:本站的技术帖子网页，遵循CC BY-SA 4.0协议，如果您需要转载，请注明本站网址或者原文地址。任何问题请咨询:yoyou2525@163.com.

相关问题 为什么训练准确率和验证准确率会波动并缓慢增加？ - why training accuracy and validation accuracy oscillate and increase very slowly? 为什么在第一个 epoch 验证准确度高于训练准确度？ - Why at first epoch validation accuracy is higher than training accuracy? 当数据集用于验证和训练时，为什么它的准确性很差？ - Why there's a bad accuracy on dataset when it's used both for validation and training? 训练准确度好但验证准确度差 - Good training accuracy but poor validation accuracy 验证精度远低于训练精度 - Validation accuracy is much less than Training accuracy 训练精度低于验证精度 - Training accuracy is less than validation accuracy 良好的训练/验证准确度，但测试准确度差 - Good training/validation accuracy but poor test accuracy 验证准确率非常低，但训练准确率很高 - Very low validation accuracy but high training accuracy 为什么训练准确率会波动？ - Why is the training accuracy fluctuating? 为什么训练时准确性和损失保持不变？ - Why is accuracy and loss staying exactly the same while training?

相关标签

粤ICP备18138465号 © 2020-2024 STACKOOM.COM