Science Forums 2

[Qfwfq]

OK guys let's try to get over it. I myself would have said that assemblers and compilers could be called translators (yup, just like from German to Russian) but if the term transducer has long been used, lets get on with the topic:

Originally Posted by SidewalkCynic

I recently learned that algorithm is a word being used without a definite defintion, as seems to be the case of a lot of terms I have been analyzing.

Anyway, I am wondering about this passage that algorithms and it seemingly exercised in fields beyond computer programming.

There is no reference, anybody able to direct me toward any?

May we then say:

• An algorithm is a procedure. One could say program although it is not necessarily a computer program. It may be defined by constructs such as a flowchart or in a language having constructs such as described in the Böhm-Jacopini theorem. This makes it a deterministic procedure even though it might incorporate making a random choice at some steps (this doesn't make the procedure itself randomly defined).
• An algorithm has the purpose of finding (one of the) solutions of a given problem problem, when executed for a given case. If it outputs a solution, it must be certain that this is a valid one. If necessary, the problem must specify a range for quantities in order to make it feasible to reach solutions.
• An algorithm can be performed on an input: a set of data which amounts to specifying a case or instance of the problem. It must also be designed to terminate within finite "time", if necessary outputting that no valid solution has been reached.
• The term sometimes gets used in ways that slacken the above two points.