Splice-site detection data
DNA data used by Meier, van de Geer, and Buelmann 2008, section 5,
and in Figure 4.4 in SLS. Data originally attributed to Burge and
Karlin (1977). See SLS for references.
There are three csv files corresponding to training, test, and
validation data.
Each has a binary response y (1 = splice site, 0 = not), and the DNA
nucleotide {a, g, c, t} at 7 positions.
splice_site_train.csv
(download): 11,220 sites, 5,110 splice site, 5,110 not.
splice_site_test.csv
(download): 93,925 sites, 4,208 splice site, 89,717 not.
splice_site_validation.csv
(download): 62,609 sites, 2,805 splice site, 59,804 not.