简体繁体 English

如何将数据从 Matlab 导入导出到 Weka 以预测类别？

[英]How to import and export data from Matlab to Weka to predict class?

原文 2016-09-21 06:17:49 8 1 matlab/ import/ machine-learning/ export/ weka

This is my proposed project: - I have developed a classifier in Weka after some experimentation.这是我提议的项目： - 经过一些实验，我在 Weka 中开发了一个分类器。 - Now I want to develop a project in Matlab which will take input attributes from users. - 现在我想在 Matlab 中开发一个项目，该项目将从用户那里获取输入属性。
- This input will be given to Weka after loading saved classification model. - 加载保存的分类模型后，此输入将提供给 Weka。 - Weka will predict the class for that instance. - Weka 将预测该实例的类别。 - Finally I want to pick this predicted class and display it on screen developed by Matlab. - 最后，我想选择这个预测类并将其显示在由 Matlab 开发的屏幕上。 Now I don't know how to import and export data from Matlab to Weka.现在我不知道如何将数据从Matlab 导入导出到Weka。 Immediate help will be appreciated.立即提供帮助将不胜感激。

1 个解决方案

Some thoughts on your problem.关于你的问题的一些想法。

As far as i know matlab has a machine learning toolbox(haven't used it).据我所知，matlab 有一个机器学习工具箱（没用过）。 So you can build your system entirely on matlab.因此，您可以完全在 matlab 上构建您的系统。
If you want to stick with Weka and use it programmatically with Java, then you can call Java methods from matlab so as to get the classification result.如果你想坚持使用 Weka 并以编程方式与 Java 一起使用，那么你可以从 matlab 中调用 Java 方法以获得分类结果。 For how to call java methods from matlab check here有关如何从 matlab 检查这里调用 java 方法
If you use Weka from GUI then you should be able to save the classification result to a file straight from GUI.如果您从 GUI 使用 Weka，那么您应该能够直接从 GUI 将分类结果保存到文件中。 Then you can load this file to matlab, retrieve the result and visualize it in matlab.然后您可以将此文件加载到 matlab，检索结果并在 matlab 中对其进行可视化。