如何在keras中可视化糟糕的预测？

Question

我正在研究文本分类问题，我想知道哪些输入我在验证时出错了预测。有没有办法在Keras做到这一点？ 我想象一下像糟糕预测的柱状图。 （X句子长度，Y数量/事件）

我发现this关于visalisation的文章，但它与我的需求有所不同。我制作的纪元很少，所以这种方法并没有为我显示任何信息。

我知道我可以save培训的历史，但据我所知

感谢您阅读我的帖子。

修改

我目前的数据框如下所示：

2018-03-27 12:36:46.933  INFO 18185 --- [er-threadpool-0] s.d.c.StandaloneAppClient$ClientEndpoint : Connecting to master spark://<spark-ip>:7077...
2018-03-27 12:36:46.989  INFO 18185 --- [pc-connection-0] o.a.s.n.client.TransportClientFactory    : Successfully created connection to /<spark-ip>:7077 after 37 ms (0 ms spent in bootstraps)
2018-03-27 12:36:47.000 ERROR 18185 --- [ rpc-client-9-1] o.a.s.network.client.TransportClient     : Failed to send RPC 7111253898393420882 to /<spark-ip>:7077: java.lang.AbstractMethodError
java.lang.AbstractMethodError: null
        at io.netty.util.ReferenceCountUtil.touch(ReferenceCountUtil.java:77) ~[netty-common-4.1.22.Final.jar!/:4.1.22.Final]
        at io.netty.channel.DefaultChannelPipeline.touch(DefaultChannelPipeline.java:116) ~[netty-transport-4.1.22.Final.jar!/:4.1.22.Final]
        at io.netty.channel.AbstractChannelHandlerContext.write(AbstractChannelHandlerContext.java:810) [netty-transport-4.1.22.Final.jar!/:4.1.22.Final]
        at io.netty.channel.AbstractChannelHandlerContext.write(AbstractChannelHandlerContext.java:723) [netty-transport-4.1.22.Final.jar!/:4.1.22.Final]
        at io.netty.handler.codec.MessageToMessageEncoder.write(MessageToMessageEncoder.java:111) ~[netty-codec-4.1.22.Final.jar!/:4.1.22.Final]
        at io.netty.channel.AbstractChannelHandlerContext.invokeWrite0(AbstractChannelHandlerContext.java:738) [netty-transport-4.1.22.Final.jar!/:4.1.22.Final]
        at io.netty.channel.AbstractChannelHandlerContext.invokeWrite(AbstractChannelHandlerContext.java:730) [netty-transport-4.1.22.Final.jar!/:4.1.22.Final]
        at io.netty.channel.AbstractChannelHandlerContext.write(AbstractChannelHandlerContext.java:816) [netty-transport-4.1.22.Final.jar!/:4.1.22.Final]
        at io.netty.channel.AbstractChannelHandlerContext.write(AbstractChannelHandlerContext.java:723) [netty-transport-4.1.22.Final.jar!/:4.1.22.Final]
        at io.netty.handler.timeout.IdleStateHandler.write(IdleStateHandler.java:302) ~[netty-handler-4.1.22.Final.jar!/:4.1.22.Final]
        at io.netty.channel.AbstractChannelHandlerContext.invokeWrite0(AbstractChannelHandlerContext.java:738) [netty-transport-4.1.22.Final.jar!/:4.1.22.Final]
        at io.netty.channel.AbstractChannelHandlerContext.invokeWrite(AbstractChannelHandlerContext.java:730) [netty-transport-4.1.22.Final.jar!/:4.1.22.Final]
        at io.netty.channel.AbstractChannelHandlerContext.access$1900(AbstractChannelHandlerContext.java:38) [netty-transport-4.1.22.Final.jar!/:4.1.22.Final]
        at io.netty.channel.AbstractChannelHandlerContext$AbstractWriteTask.write(AbstractChannelHandlerContext.java:1081) ~[netty-transport-4.1.22.Final.jar!/:4.1.22.Final]
        at io.netty.channel.AbstractChannelHandlerContext$WriteAndFlushTask.write(AbstractChannelHandlerContext.java:1128) ~[netty-transport-4.1.22.Final.jar!/:4.1.22.Final]
        at io.netty.channel.AbstractChannelHandlerContext$AbstractWriteTask.run(AbstractChannelHandlerContext.java:1070) ~[netty-transport-4.1.22.Final.jar!/:4.1.22.Final]
        at io.netty.util.concurrent.AbstractEventExecutor.safeExecute(AbstractEventExecutor.java:163) ~[netty-common-4.1.22.Final.jar!/:4.1.22.Final]
        at io.netty.util.concurrent.SingleThreadEventExecutor.runAllTasks(SingleThreadEventExecutor.java:404) ~[netty-common-4.1.22.Final.jar!/:4.1.22.Final]
        at io.netty.channel.nio.NioEventLoop.run(NioEventLoop.java:463) ~[netty-transport-4.1.22.Final.jar!/:4.1.22.Final]
        at io.netty.util.concurrent.SingleThreadEventExecutor$5.run(SingleThreadEventExecutor.java:886) ~[netty-common-4.1.22.Final.jar!/:4.1.22.Final]
        at io.netty.util.concurrent.FastThreadLocalRunnable.run(FastThreadLocalRunnable.java:30) ~[netty-common-4.1.22.Final.jar!/:4.1.22.Final]
        at java.lang.Thread.run(Thread.java:748) ~[na:1.8.0_161]

在确认之后，我想像pythonic833那样得到类似的内容：

+----------------+----------+
|    sentence    | language |
+----------------+----------+
| test sentence. | english  |
+----------------+----------+

我可以在培训期间实现这一点（我有单独的验证数据集），还是我应该使用+----------------+----------+-----------+ | sentence | language | validated | +----------------+----------+-----------+ | test sentence. | english | english | +----------------+----------+-----------+ 并自己构建它？