The Communities in a network are detected by investigating the corresponding graph and finding dense clusters of vertices. The decoding algorithm SASH determines the initial codeword of communities that would most likely result in specific clusters. SASH checks various candidate codewords at clustering types until the codeword with the smallest discrepancy from the observed dataset is located. Using the dataset Zachary’s Karate Club, errors within the algorithm that lead to a significantly lowered accuracy from expectations will be highlighted, as well as potential ways that could amend the issues to optimize performance.
