2, half of the insertion order: because it has been determined that the first part of the ordered sequence, so in the search for the insertion of the location of the search can be used to find half of the method of searching, improve efficiency.
3, Hill sort: also known as reduced incremental sorting method. The sequence to be sorted into a number of smaller sub-sequence, and then one by one using the direct insertion sorting method of sorting, and then finally a more orderly sequence for a sort, mainly to reduce the number of moves, improve efficiency. The principle should be from the unordered to gradually ordered, than directly from the unordered to the ordered move will be less.
4, bubble sort: repeatedly scanning the sequence to be sorted, in the scanning process in turn to compare the size of the two neighboring elements, if the reverse order on the exchange of position. The first trip, starting from the first data, comparing the two adjacent data, (in ascending order, for example) if large on the exchange, to get a maximum of data at the end; and then the second trip, only scanning the first n-1 elements, get the next largest placed in the penultimate position. And so on, and finally get the ascending sequence. If during the scanning process, it is found that there is no exchange, indicating that the sequence has been arranged, directly terminate the scan. So at most n-1 scans are performed.
5, fast sort: Idea: bubble sort can only eliminate a reverse order at a time, in order to eliminate more than one reverse order at a time, the use of fast sort. A keyword as an axis, from left to right in order to compare with it, and then exchanged, the end of the first trip, the sequence can be divided into two sub-sequence, and then segmented for quick sorting, to achieve high efficiency.
In addition there are selection, merging, allocation sorting, etc. and their subclasses of sorting