Explore-Data-via-Weka

Ref: https://www.cs.waikato.ac.nz/ml/weka/courses.html

Data: weather.nominal

Open file weather.nominal.arff

Use a filter to remove an attribute

• Check the filters
  • supervised vs unsupervised
  • attribute vs instance
• Choose the unsupervised attribute filter Remove
• Check the More information; look at the options
• Set attributeIndices to 3 and click OK
• Apply the filter
• Recall that you can Save the result
• Press Undo

Exercise -1

1. Remove instances where humidity is high.

Exercise -2

1. Download the data "trains-transformed.data" from https://archive.ics.uci.edu/ml/datasets/Trains;
2. Create an ''Arff data'' version for Weka;
3. Check whether it has missing values? If so, remove the instances that contain the missing values.