Loading and Exploring Data in WEKA

Step 1: Start WEKA

• Launch WEKA by double-clicking the WEKA icon or executing the command.

Step 2: Open the Explorer

• Click on "Explorer" (the first option in the GUI Chooser).

Step 3: Load a Dataset

• In the Explorer window, click the "Open file" button.
• Navigate to the folder where your dataset is located (e.g., weather.arff or any .arff, .csv file).
• Select the file and click Open.

WEKA supports formats like:

.arff (Attribute-Relation File Format – native to WEKA)

.csv (Comma-Separated Values)

.xrff (XML-based ARFF)

.json, .bsi (in newer versions)

Step 4: View Dataset Summary

• After loading, you will see:
  • The relation name (dataset title).
  • Number of instances (rows of data).
  • Number of attributes (columns or features).
  • A list of attributes on the left.

Step 5: Explore Individual Attributes

• Click on an attribute from the list on the left.
  • The right pane shows:
  • Histogram for the selected attribute.
  • Frequency count for categorical attributes.
  • Mean, StdDev, Min, Max for numeric attributes.

You can visually inspect class balance, missing values, and value distributions here.

Step 6: Explore Data Preprocessing Options (Optional)

• Use the "Filter" option to:
  • Normalize or standardize data
  • Replace missing values
  • Convert nominal to numeric attributes
  • Remove or select specific attributes
  • Click Choose under "Filter" to apply a transformation.

Step 7: Save the Modified Dataset (Optional)

Click Save to export the preprocessed dataset to .arff or .csv.