more detailed information in ITALIAN: REPORT.MD
Comparing K Nearest Neighbor with different k and metrics.
- pandas: load and manipulation of DataSet
- sklearn: algo, stats and splitting of TrainingSet
note: see
main-from-scratch.py
+lib.py
for a an implementation without external library
main.py [dataset] [test_size]
- dataset: [wine|breast-cancer|letters|poker] - select DataSet. Default is wine
- testset_size: [0,1] - TestSet dimension
The DataSet taken from archive.ics.uci.edu are: