Archaeological evidence suggests that, as early as 7,000 years ago, human skulls were drilled for healing purposes. Over time, understanding the brain has become a priority for science. Neuroscience, for example, is dedicated to exploring how the brain functions, from the molecular level to its global structure, and how these processes give rise to behavior, emotions, perception, and cognition. Inspired by these discoveries, the field of machine learning has developed techniques such as artificial neural networks, which simulate, at a basic level, the way the human brain processes information. One of the central challenges of neuroscience is understanding which regions of the brain are activated during situations involving social behavior. To address this issue, observing videos of mice interacting is often used as a method to identify and characterize social behaviors in these animals. Understanding the temporal evolution of behaviors can provide access to how animals process information. In this work, a miniscope is used to obtain calcium traces from neurons, which are analyzed by video through luminescence data, from which it is possible to use this data as input to train a machine learning model. In addition, we present preliminary results connected with recurrence analysis and classification processes.