Looking for patterns in data to try to separate them into different groups by using, for example, similarity or dissimilarity rule is a common problem solved through clustering. This could be approached as a mathematical programming model that minimizes the distance between the data inside a cluster thus each will be as more homogeneous as possible. In this work, we use a clustering model for the k-means problem-based in mathematical programming. The computational test was conducted on simulate data for a bovine animal database containing physical and breed characteristics of each animal but with unknown patterns. The interest is to identify patterns among the different animals by grouping them from the solutions of an integer linear optimization model. The computational results are studied from a set of descriptive statistical procedures to validate this research.