User:JamesRelativeCoder1980/sandbox

Entropy Balanced Search

The idea behind entropy balanced search, is simply to be able to take the frequencies of the symbols within an array (or other structure, such as a doubly linked list) and be able to insert those items in such a way into a table of containers in which the entropy (frequency in this case) is used to determine what "index" that this item will be found in if you were to look for this item. The idea of this concept sounds simple, and really is a matter of determining how many containers you wish to have, the amount of different symbol types, and balancing based on simple frequency analysis that would be generally served by a rotating insertion method in which if just by "frequency rank" in an collection size of 1000, values of a range that were defined to be of 4000 different elements, then such 0, 3999, 2999 and 1999 would be in the same collection, etc...

This is a concept that myself, James Bernard Schumacher III, have come up with and have invested some time in my own research, and will continue to edit this article with more specific information than just this general description in the future.

However, keep in mind that if you have the frequencies of a certain symbol, you have jumped to that location already... So this is useful only if you have different data structures referencing this data, and in particular when related to compression there's a copyright on a method what James Bernard Schumacher III calls "fast node technology".