Probabilistic Assumptions

The CLICK algorithm makes the following assumptions:

1.

Similarity values between mates are normally distributed with mean $\mu_T$ and variance  $\sigma_T^2$.

2.

Similarity values between non-mates are normally distributed with mean $\mu_F$ and variance  $\sigma_F^2$.

3.

$\mu_T > \mu_F$

These assumptions are justified both empirically and theoretically by the Central Limit Theorem.