Question $2$

A particular biometric system generates a score $[$S$](0S14,$ which measures the

similarity between a test sample and a corresponding reference model for claimed

identity. An identification decision is then made according to whether $S>$ or

where $$ is a variable threshold value $(114).$ The FRR $(\%)$ and FAR $(\%)$ for this

system for a range of values of $$ is determined experimentally, with the results shown

in Table $1.$

$($i$)$ Graphically determine the system's equal error rate $($EER$).$

$($ii$)$ Suggest and justify a threshold setting which might be suitable when using the system to control physical access to a University Teaching Laboratory.

$\backslash$table$[[\backslash$theta $,$FRR$,$FAR$],[1,1\mathrm{.}3,23\mathrm{.}0],[2,1\mathrm{.}4,14\mathrm{.}6],[3,1\mathrm{.}6,9\mathrm{.}0],[4,1\mathrm{.}8,6\mathrm{.}2],[5,2\mathrm{.}0,4\mathrm{.}6],[6,2\mathrm{.}2,3\mathrm{.}4],[7,2\mathrm{.}6,2\mathrm{.}8],[8,3\mathrm{.}0,2\mathrm{.}4],[9,3\mathrm{.}6,2\mathrm{.}0],[10,4\mathrm{.}5,1\mathrm{.}8],[11,5\mathrm{.}4,1\mathrm{.}7],[12,6\mathrm{.}4,1\mathrm{.}5],[13,7\mathrm{.}5,1\mathrm{.}3],[14,8\mathrm{.}6,1\mathrm{.}0]]$