Abstract

Artificial neural networks (ANN) have been developed as generalizations of mathematical models of biological nervous systems. Generalized regression (GR) and multiple linear regression (MLR) models for shelf life prediction of dairy product were developed. Results of both the models were evaluated with three types of prediction performance measures viz., mean square error, root mean square error and coefficient of determination R², and compared with each other. Based on these results, regression equations were developed. From the study, it is concluded that ANN and statistical computerized methods can be employed for predicting shelf life of dairy products.

Key words: Artificial neural networks (ANN), shelf life, dairy product, prediction, generalized regression, multiple linear regression.