accuracy vs. reliability
Maybe I've gotten myself in hot water here. For me, accuracy refers to the closeness to the truth, for example as reflected by the number of digits after the decimal. More the digits after the decimal, the closer to the truth you might be. Reliability refers to consistency. It may not be very accurate, but it may be consistently good enough that you can make a decision based on it. For example, saying "go right at the split in the road" is as reliable as it can get, but has no concept of accuracy of measurement in it.
But then, I hope no other monk pounces on me for saying the above as I may be spewing more than I have chewed.
(for that matter, I myself don't understand the difference between 'accuracy' and 'precision', but that is neither here nor there for the problem in question).
| [reply] |
| [reply] |
So, you're saying it's ok if the algorithm's output is wrong, as long as it's consistently wrong for the same inputs?
Put that way, no, that is not what I am saying (although I knew it would come and bite me in my butt ;-)
Otoh, I might be saying that _if_ the output is wrong, it would be nice to know that it is, and to know by how much it is.
Other than that, of course, I would like it to be good, fast, cheap, free, in Perl, accurate, precise, reliable, concise, and free of carbs.
| [reply] |