4

I was currently doing a project on Vehicle classification and it has almost finished now but I have several confusion about the plots I get from my Neural Network

I used 230 images [90=Hatchbacks,90=Sedans,50=SUVs] for classification on 80 feature points. Thus my vInput was a [80x230] matrix and my vTarget was [3x230] matrix

Classifier works well but I don't understand these plots or if they are abnormal or not.

My neural Network
Neural Network

Then I clicked these 4 plots in the PLOT section and got these sequentially.

Performance Plot
Performance
Training State
training state
Confusion Plot
Confusion plot
Receiver Operating Characteristic Plot
ROC plot


I know the images they are a lots of images but I know nothing about them. On the matlab documentation they just train the system and plot the graph
So please someone briefly explain them to me or show me some good links to learn them.

4

1 回答 1

6

前两个图显示了训练统计信息。

性能图以对数刻度显示所有数据集的均方误差动态。训练 MSE 总是在减少，因此您应该对它的验证和测试 MSE 感兴趣。您的绘图显示了完美的训练。

Training State向您显示其他一些训练统计数据。

梯度是在对数尺度上每次迭代的反向传播梯度值。5e-7意味着您达到了目标函数的局部最小值的底部。

验证失败是验证 MSE 增加其值时的迭代。很多失败意味着训练不足，但在你的情况下它还可以。Matlab 连续 6 次失败后自动停止训练。

另外两个图显示了训练后网络模拟的结果。

混淆情节。在您的情况下，它是 100% 准确的。绿色单元格代表正确答案，红色单元格代表所有类型的错误答案。

例如，您可以将第一个（训练集）解读为：“来自第 1 类的 59 个样本被正确分类为第 1 类，来自第 2 类的 13 个样本被正确分类为第 2 类，来自第 3 类的 6 个样本被正确分类为归为第 3 类”。

接收者操作特征图显示了同样的事情，但以不同的方式 - 使用ROC 曲线：

在此处输入图像描述

于 2013-11-17T00:05:00.730 回答