"false accept", "false reject"

[Mark Mandel]

These are NPs in biometrics and related fields. Reading SpeechTech
magazine online, I encountered (http://tinyurl.com/d3c5hq)
  This results in a reduction of matchable data points, which in turn
increases the incidence of false rejects.

Raw ghits:
about 6,560 for "false rejects"
about 18,400 for "false reject"
about 3,610 for "false accepts"
about 20,600 for "false accept"

Judging from the first 20 of each, almost of these are genuine.

Neither noun (in this sense) is in OED Online or Merriam-Webster
Online. OED has only the familiar sense of reject (n.) --

    1.    a. One who is rejected; a castaway. Obs.
    b. One who is rejected or discarded by others, esp. as unsuitable
for some activity (orig. for military service).
    2. A thing rejected as unsatisfactory.
    3. attrib.    a. Appositive.
    b. General attrib. uses.

-- and does not have accept (n.) in any sense.

I suspect that both Ns are stressed on the final syllable, unlike the
present-day pronunciation of the familiar noun "RE-ject".

Mark Mandel

------------------------------------------------------------
The American Dialect Society - http://www.americandialect.org