| ******************************************************* | *************************************************** | ********************************************** | ****************************************** | T H E M O N K' S P R O B L E M S | ****************************************** | ********************************************** | *************************************************** | ******************************************************* | | | | The MONK's problems are a collection of three binary classification | problems over a six-attribute discrete domain. Each training/test data | is of the form | | : -> | | where is an ASCII-string, represents the value of | attribute # n, and is either 0 or 1, depending on the class | this example belongs to. The attributes may take the following values: | | attribute#1 : {1, 2, 3} | attribute#2 : {1, 2, 3} | attribute#3 : {1, 2} | attribute#4 : {1, 2, 3} | attribute#5 : {1, 2, 3, 4} | attribute#6 : {1, 2} | | Thus, the six attributes span a space of 432=3x3x2x3x4x2 examples. | | | | /*********************************************************************\ | *********************************************************************** | \*********************************************************************/ | | | | The "true" concepts underlying each MONK's problem are given by: | | MONK-1: (attribute_1 = attribute_2) or (attribute_5 = 1) | | MONK-2: (attribute_n = 1) for EXACTLY TWO choices of n (in {1,2,...,6}) | | MONK-3: (attribute_5 = 3 and attribute_4 = 1) or | (attribute_5 != 4 and attribute_2 != 3) | | (with "!=" denoting inequality). | MONK-3 has 5% additional noise (misclassifications) in the training set. | | | | /*********************************************************************\ | *********************************************************************** | \*********************************************************************/ 0, 1 |classes attribute#1 : 1, 2, 3 attribute#2 : 1, 2, 3 attribute#3 : 1, 2 attribute#4 : 1, 2, 3 attribute#5 : 1, 2, 3, 4 attribute#6 : 1, 2