A clustering method to identify representative financial ratios

When companies evaluate their performance, it is impractical to take all of their financial ratios into consideration. To evaluate the financial performance of a company, only a fraction of the available financial ratios are considered and selected as evaluation criteria. In general, financial ratios presented as sequences (or called financial ratio sequences), are first clustered and then a representative indicator is chosen from each cluster to serve as an evaluation criterion. To cluster financial ratios, we propose a clustering method in which the financial ratios of different companies with similar variations are partitioned into the same cluster. In other words, a fuzzy relation is proposed to represent the similarity between the financial ratios, and a cluster validation index is also provided to determine the number of clusters. Once the financial ratios are clustered, the representative indicator for each cluster will be identified.