The main problem for pig producers is to know how much feed must consume an individual pig to maximize the conversion of feed into live weight over time. Precision livestock farming with the use of feeding robots allow farmers and researchers to analyze the variability of variables like live weight, average daily gain, feed intake, number of meals, time spent eating over the fattening period. In this research we addressed the interrelationships of these variables to explore how they affect pig growth and fattening process. Given the expected degrees of correlation among several of these variables, we use principal component analysis to reduce dimension and retain the most significant ones in view of facilitating the identification of different feeding behaviors or growth typologies. For the principal components analysis, we consider a batch of fattening pigs with individual measurements regarding feeding behavior and growth provided by the “Centre d’estudis porcins” (Torrelameu, Spain). After that, a multiple linear regression model is proposed to explain feed conversion rate in terms of the most relevant variables present in the main principal components. We tested the model against a new batch of fattening pigs and we found that our model fits the data, with an adjusted R-squared value of 95%. Results allowed us to identify three feeding behaviors and open the door for a more precise, tailored and performant feeding strategies of growing pigs.