7
$\begingroup$

No apologies: I have not attempted to research this (beyond reviewing the list of questions CV provided that may have answered this query). I taught this in class last week for diagnosing logistic multiple regression models, and I warned the students in advance that I did not know the origins of the name.

What is the history of the name of the ROC curve: Receiver Operating Characteristic?

I recall something about it being mentioned in an avocational book (like The Lady Tasting Tea or one of Mario Livio's books)...but if anyone has some history to share, that would be welcome.

$\endgroup$
  • 8
    $\begingroup$ There's a section in Wikipedia with links to citations and all right here. The "receiver" was picking up enemy radars and trying to intercept communication. Operating characteristics is literally "how well it works". $\endgroup$ – AdamO Apr 17 '18 at 13:40
  • $\begingroup$ Very useful link...now I wonder what the references are that first connect the signal detection field to medical literature. $\endgroup$ – Gregg H Apr 17 '18 at 13:51
  • 2
    $\begingroup$ you can read link 34 $\endgroup$ – AdamO Apr 17 '18 at 13:52
13
$\begingroup$

The earliest book reference that I know of is

Woodward, P. M. (1953). Probability and information theory with applications to radar. London: Pergamon Press.

but the concept, which was developed during World War II for the analysis of radar receivers, might have been published earlier than 1953 in journal articles (after the War was over) or in the multivolume series of texts published by the MIT Radiation Laboratory about their research during World War II.

$\endgroup$
2
$\begingroup$

Earliest article I can find is from 1954:

  • Peterson, W., Birdsall, T., Fox, W. (1954). The theory of signal detectability, Transactions of the IRE Professional Group on Information Theory, 4, 4, pp. 171 - 212.

Abstract:

An optimum observer required to give a yes or no answer simply chooses an operating level and concludes that the receiver input arose from signal plus noise only when this level is exceeded by the output of his likelihood ratio receiver. Associated with each such operating level are conditional probabilities that the answer is a false alarm and the conditional probability of detection. Graphs of these quantities called receiver operating characteristic, or ROC, curves are convenient for evaluating a receiver. If the detection problem is changed by varying, for example, the signal power, then a family of ROC curves is generated. Such things as betting curves can easily be obtained from such a family.

$\endgroup$
  • 1
    $\begingroup$ Note that this is after the book listed in the existing answer. It could be the first article, though. $\endgroup$ – gung Mar 18 at 15:28

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.