Pattern recognition of coals using the two-phase model

Principal components analysis and cluster analysis were used to the study o f several American and Romanian coals using data obtained by applying the t wo-phase model of coal. A two-component model, which explained 95.80 % of t he total variance was retained. Scores plots appeared to be very useful as a display tool for examining relationships between coals, looking for trend s, groupings and outliers. This "chart of similarity" may be successfully u sed, for instance, in the studying origin and quality of coals.