Prediction of the composition and percentage of ingredients in a functional blend for yoghurt production


Identifying the components in a blend and knowing the exact percentage of each one is essential to ensure its correct labeling, avoiding possible consumer complaints. In this case study we explain how the Ingredient Analysis tool within Chemometric Brain’s software analyzes thousands of data available in our ingredients database and is able to predict the composition of any food sample.


A dairy product company specialized in manufacturing different types of yoghurt products was interested in determining the composition of a powder blend of stabilizers used in their recipes. By implementing Chemometric Brain’s analysis of ingredients, the company could determine the composition and the percentages of any ingredient in the stabilizer blend, ensuring the correct labelling in the final product.  


A stabilizers blend for yoghurts used in a dairy company.


Determining the composition and percentage of all ingredients in a blend of stabilizers for yoghurts.


To determine the composition and percentage of all different ingredients, our customer benefited from another functionality available in Chemometric Brain: the ingredients analysis tool. Figure 1 shows the results obtained for a new batch of yoghurt stabilizer blend received in the company. In this case, the system performs a learning process by analysing the thousands of data available in our ingredients database, their percentage in different products and their response to NIR radiation. After this process, the software is prepared to predict the composition of any spectrum. Once the data training has been carried out, the prediction process starts.

The software starts working on the prediction of the composition of the selected sample, which will be displayed on the screen as soon as the process is finished.

The software shows not only the components but also the percentage in which each of them is in the product.


By acquiring the ingredients Analysis tool offered by Chemometric Brain Software-as-a-Service, customers will be able to predict both the components that are part of a blend and the proportion in which they are present. This methodology can be successfully applied not only in the identification of macrocomponents but also in the identification of microcomponents and thereby insure that you get the same product every time and avoid food fraud.