gptsne

Rough usage (from the src/ directory):
python3 -m gptsne.gptsne_mo --help
e.g. python3 -m gptsne.gptsne_mo -d COIL20 --dir "datasets/"

• Datasets used in the paper are in datasets/
• Add your own datasets in csv format, with a header line:
  Header: classPosition,#features,#classes,seperator. e.g.
  classLast,1024,20,comma (from COIL20.data)
• Most GP parameters are configured in gpptsne/tsnedata.py