Discrimination of adulteration orange juice by Linear Discriminant Analysis (LDA)

In this paper the Linear Discriminant Analysis (LDA) techniques is used to discriminate various orange juice samples [1] of pure juice, pure juice adulterated with sugar, tap water etc. Orange juices are a popular refreshment beverage and nutrition source. However, the adulteration of pure juice has now become a common practice. Adulteration ranges in sophistication from simply diluting juice with water to adding sugar, the adulterant that is most difficult to detect. Introducing these ingredients is not illegal; however, knowingly selling the resulting product as pure juice constitutes fraud. Here, the sample orange juice samples (pulpy orange) has been taken from coca cola brand. By forming samples phase data into matrix form at different frequencies in Constant phase (property of electrode-electrolyte interface of Polymer coated Copper probe dipped in ionic solution) region as well as data taken at different time. Euclidean distance from any unknown sample with the given known samples has been calculated using LDA algorithm. The sample with shortest distance will show the best maximal characteristics of the unknown sample. The MATLAB software has been used for the whole calculation and discrimination plotting of data.