Abstract
A noninvasive technique using a low-cost portable microwave, also known as a microwave system (MiS) has been developed to measure the fat depth. A multiple linear regression model has been built together with the partial least square model to build a predictive model for better estimation of the c-site fat depth of organic animals. Both models use fivefold cross-validation using R-squared and other accuracy measures as indicators of precision. Since multicollinearity is a major problem due to high-dimensional data, K-means, and principal component analysis have been used to reduce the explanatory variables. In short, the partial least square model performed much better even though the accuracy measures are similar to both models, as multicollinearity and assumptions are not satisfied for the multiple linear regression model.