covertype (Q6033070)

**Author**: Jock A. Blackard, Dr. Denis J. Dean, Dr. Charles W. Anderson \N**Source**: [LibSVM repository](http://www.csie.ntu.edu.tw/~cjlin/libsvmtools/datasets/) - 2013-11-14 \N**Please cite**: For the binarization: R. Collobert, S. Bengio, and Y. Bengio. A parallel mixture of SVMs for very large scale problems. Neural Computation, 14(05):1105-1114, 2002.\N\NThis is the famous covertype dataset in its binary version, retrieved 2013-11-13 from the libSVM site (called covtype.binary there). Additional to the preprocessing done there (see LibSVM site for details), this dataset was created as follows:\N-load covertpype dataset, unscaled.\N-normalize each file columnwise according to the following rules:\N-If a column only contains one value (constant feature), it will set to zero and thus removed by sparsity.\N-If a column contains two values (binary feature), the value occuring more often will be set to zero, the other to one.\N-If a column contains more than two values (multinary/real feature), the column is divided by its std deviation.\N-duplicate lines were finally removed.\N\NPreprocessing: Transform from multiclass into binary class.