Last Updated on 2021-08-31 by Clay
In the past I have no any experience of Weka, but in the meeting of fate, I came into contact with Weka that a tool suitable for data mining and machine learning, so I simply recorded how to install it here.
This article is divided into two subsections. The first section is how to install the Weka tool, and the second section is to test whether the Weka I installed can work normally.
First I need to introduce about Weka.
What is Weka?
As mentioned above, Weka is a software that convers data mining, machine learning and other fields. It is a tool developed by Waikato University in New Zealand using Java. The four characters Weka are also "Waikato Environment for Knowledge Analysis" abbreviation.
I heard that Weka is also the name of the bird on the icon.
Compared with most tools used in the field of data exploration and machine learning, Weka has a graphical interface. Users can select the data they want for processing (for example, classification) without writing any code.
It is quite convenient.
Download Weka
Weka is an open source software and it can executed on Windows, Mac OS and Linux system. So you just need to go the the following link and select the OS you used: https://waikato.github.io/weka-wiki/downloading_weka/
In here, I select the Linux system version, unzip it and use the following command to execute.
cd weka-3-8-4/
sh weka.shthen you can see the Weka window opened.
Use Decision Tree To classify
In order to make sure that the downloaded Weka can work normally, you can choose to load the weather data built in Weka and use the decision tree for simple classification, and try to see if Weka cna successfully classify.
Step 1. Open Explorer
Step 2. Load Data
Select Open file ... > weka-3-8-4/data/weather.nominal.arff > Open, and load weather data.
You can see the following data:
Select the Edit button to view the original data:
Step 3. Classify
Switch to Classify and select the Choose button to select the classifier.
You can choose J48 that is a binary classification decision tree.
Press Start button and you will see the results of cross-validation (10 folds).
If the data is displayed normally, it means that the Weka tool can operate normally.