Classification DataSet

• Reuters
  • RCV1 (Reuters Corpus Volume 1)
    • RCV1: A New Benchmark Collection for Text Categorization Research (2004)
      • pdf
    • Reuters Corpus
      • RCV1-v2/LYRL2004: The LYRL2004 Distribution of the RCV1-v2 Text Categorization Test Collection
        • lyrl2004_vectors_test_pt0.dat.gz : 159879168
        • lyrl2004_vectors_test_pt1.dat.gz : 161878016
        • lyrl2004_vectors_test_pt2.dat.gz : 158580736
        • lyrl2004_vectors_test_pt3.dat.gz  : 149512192
        • lyrl2004_vectors_train.dat.gz : 18620416
• LIBSVM data sets
  • Data: Classification (Multi-class)
    • heart
    • glass UCI
    • iris
    • Connect-4
    • wine
    • news20
      • Source: [KL95a]
      • Preprocessing: First 80/20 training/testing split.Also see this page [JR01a]
      • # of classes: 20
      • # of data: 15,935 / 3,993 (testing)
      • # of features: 62,061 / 62,060 (testing)
      • news20.scale.bz2 (scaled to binary encoding; then unit length for each instance)
      • news20.t.scale.bz2 (testing) (scaled to binary encoding; then unit length for each instance)
        • ./svm-train news20.scale
        • ./svm-predict news20.t.scale  news20.scale.model news20.t.predict
• OHSUMED Test Collection
• UCI
  • UCI Knowledge Discovery in Databases Archive
  • Classification
• 20 Newsgroups
  • 20news-bydate.tar.gz - 20 Newsgroups sorted by date; duplicates and some headers removed (18846 documents)