Biclustering Swiss In-home Food Consumption Across Consumer Groups and Foods

This study is the first to apply biclustering to households’ home consumption of food and finds patterns of households and foods that may account for factors of consumption in Switzerland. Leveraging biclustering theory and rich data from the Swiss Federal Statistical Office, we apply the FABIA algorithm to define and label five overlapping biclusters - basic, vegan processed, old & poor, vegan raw, and sweet-toothed - which are distinguished mainly by age and household size. Our results may help to target food and nutrition policy and marketing activities in the future by using clearer behavioural information about consumers. Vegan behaviour can also be stimulated using evidence that their nutrition and lifestyle are sustainable.