There are a variety of classification algorithms and each classification has its own reasons. One way to classify the types of algorithm is to consider the paradigms and methods used to design the algorithm. Some of the paradigm used in formulating an algorithm will be presented in this section. Each of these paradigms can be used in many different algorithms.
1) Divide and Conquer, a paradigm to divide large problems into problems of smaller ones. The division of this issue carried out continuously until it was the small problem that is easy to solve. In summary solve the whole problem by dividing a large problem and then solve the small problems that are formed.
2) Dynamic programming, dynamic programming paradigm would be appropriate if used on a problem that contains a sub-optimal structure (and contains some of the issues that overlap. Paradigm glance looks similar to the paradigm of Divide and Conquer, both trying to divide problems into smaller sub-problems, but there are intrinsic differences of character of the problems faced.
3) Method of greed. A greedy algorithm is similar to a dynamic programming, the difference between the answers of the subproblems do not need to know in each stage, and using the "greedy" what is seen best at the time.
