Rows: Vienna variants (mixed of meat)
Cols: Floury, Spicy, Rubbery, White Meat VS Dark Meat
Interpret: Component 1 is explaining “the difference within” Vienna variants
Rows: Pavo variants (Turkey)
Cols: Rubbery & White Meat VS Acidic, Bitter, Floury, Dark Meat
Interpret: Component 1 is explaining “the difference within” Pavo variants
We see that DiCA, especially in Fixed Data showed that it could classify the observations in a Unsupervised manner with relatively high accuracy (82%). However, comparing to BADA, DiCA has inferior classification performance (only 48% accurate on Random Data)
In DiCA, we essentially performed CA on the group means of the dataset after it is binarize (categorize). The grouping strategy that we have chosen derived from our findings in PCA regarding the differentiation between Pavo and Viena sausages. Overall the findings were simialar to that in PCA and BADA regarding observations on each component 1 and 2. One point of interest in DiCA is that it can also point out the eaxct bins of each variables that stood out the most. Knowing this level of details would allow us to form hypothesis about the fairness of ratings based on their distribution within bins. This point of thoughts also lead us to suspect more about the assumption that Panelists rated similarly as presented in PCA.