提交 2559869e 编写于 作者: happyboxman2007's avatar happyboxman2007

Update Basic classification: Classify images of clothing.md

上级 cfed9e2a
......@@ -83,4 +83,4 @@ test_loss, test_acc = model.evaluate(test_images, test_labels, verbose=2)
如上图所示,train_acc=0.91,test_acc=0.88,0.88 < 0.91 并且差距不小,造成这个的原因是过拟合,意思就是模型在训练集上的表现比测试集上的好,有一部分原因是模型记住了训练集的数据和其中的noise,从而对新数据的预测造成负面影响
代码地址: https://codechina.csdn.net/csdn_codechina/enterprise_technology/-/blob/master/Basic%20classification:%20Classify%20images%20of%20clothing.md
\ No newline at end of file
代码地址: https://codechina.csdn.net/csdn_codechina/enterprise_technology/-/blob/master/Basic%20classification:%20Classify%20images%20of%20clothing.md
Markdown is supported
0% .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑!
想要评论请 注册