The excessive and inappropriate use of antibacterials has led to the development of antimicrobial resistance, posing a significant threat to human and animal health. To address this challenge, the use of biosensors and advanced Machine Learning techniques holds the promise of rapid and accurate identification of microorganisms, such as bacteria, in substance samples. This study focused on the application of Machine Learning to data obtained by electrical impedance spectroscopy using a biosensor – electronic tongue – containing no specific receptors to identify the presence of six bacteria commonly found in cow’s milk. The results revealed that the Long Short-term Memory Neural Network model reached the best results, with an average accuracy of 93.46% and satisfactory results in all metrics evaluated. Furthermore, the Random Forest model also stood out as a robust alternative, presenting balanced performance. The integration of electronic tongues with Machine Learning models to bacterial identification leads to rapid and efficient diagnoses, with potential application in clinical and laboratory practices.