A list consisting of Training and Test data along with corresponding class labels.
KOS_DataA list consisting of:
(179 x 4) Matrix of training data features. the first two features satisfy sqrt(x_i1^2 + x_i2^2) > 2/3 if the ith sample is in class 1. Otherwise, they satisfy sqrt(x_i1^2 + x_i2^2) < 2/3 - 1/10 if the ith sample is in class 2. The third and fourth features are generated as independent N(0, 1/2) noise.
(94 x 4) Matrix of test data features. the first two features satisfy sqrt(x_i1^2 + x_i2^2) > 2/3 if the ith sample is in class 1. Otherwise, they satisfy sqrt(x_i1^2 + x_i2^2) < 2/3 - 1/10 if the ith sample is in class 2. The third and fourth features are generated as independent N(0, 1/2) noise.
(179 x 1) Vector of class labels for the training data.
(94 x 1) Vector of class labels for the test data.
Lapanowski, Alexander F., and Gaynanova, Irina. ``Sparse Feature Selection in Kernel Discriminant Analysis via Optimal Scoring'', preprint.