some practice for decision tree

• the tree.py can be used in other applications.
• train.txt include the training data.
• training file:
  • the first line should be the list of attributes that do not include the target attribute.
  • the second line should be the target attribute.
  • from third line, there is no empty line in this file.
  • put the training data in order above.
• cross_chain file:
  • the first line is the list of attributes that do not include the target attribute.
  • the string in the "()" is the real correct class name.
• the theory of this code is from the chapter 3 of Machine learning written by Tom M.Mitchell.