Classification of orange juice adulteration using LDA, PCA and ANN

This paper describes the classification of orange juice samples using Linear Discriminant Analysis (LDA), Principal Component Analysis (PCA) and Artificial Neural Network (ANN) detection techniques. The performance of the fractional order capacitor type sensor to detect different types of adulterant in orange juice is also reported here. Comparison between PCA, LDA and ANN detection techniques are done to discriminate various orange juice samples of pure juice, pure juice adulterated with sugar, tap water, etc. Here, the sample orange juice has been taken from coca cola brand. The sample data is phase data at different frequencies in Constant phase region as well as data taken at different time. Here Phase data is arranged into matrix form for data analysis. The data classification with PCA, LDA and ANN are compared using Discrimination index (DI) and finally the best classification method has been identified using DI value for the samples of orange juice.