Principle of Mining Co-Location Patterns

Chapter 3 introduced the fundamental theory of decision tree generation including tree construction, algorithm modeling, and attributes splitting criteria, pruning, and accuracy evaluation of decision tree performance. This chapter will present an innovative method called a co-location spatial decision tree induction, which incorporates co-location (also called co-occurrence) mining into the decision tree. This chapter will describe the details of the co-location decision tree construction, algorithms, modeling, decision rule, node splitting criterion, node merging criterion, and leaf stopping criteria and then will give an example for illustrating the calculation process of the co-location spatial decision tree induction algorithm.