Probabilistic Approach

Another approach is to build a probabilistic model of contamination errors and try to devise an algorithm which given C(H) reconstructs the original clique graph H w.h.p. (with high probability).

One of the simplest probabilistic models for contamination errors is a random corrupted clique graph. The contamination errors are represented by flipping a fraction p < 0.5 of the edges of the original clique graph H. This can be thought of as flipping edges with probability p < 0.5. We will denote by $\Omega(H,p)$ the set of all corrupted clique graphs derived from H with contamination error fraction p using this model.