- The targets should be self-explaining
- In all files, the last column is the target itself
- leave that one out for the experiments (it's just fixed parameter of the library which produced the data)
- For education, it still shows the numerical value 'education-num' (1-16) instead of the more useful categorical description (pre-school - PhD)
- I will change that in the coming days, but need to manually write a generalization hierarchy file for it...
cassinius/akhci-2017
Sample data for the student groups of the AK-HCI 2017 course on interactive ML for anonymization
JavaScriptApache-2.0