Early detection of antibiotic resistance is crucial, especially for vulnerable patients or those on pro- longed treatments with a single antibiotic. Re- searchers are employing advanced technologies such as Machine Learning to tackle this issue. Researchers have used MALDI-TOF mass spec- trometry to predict antibiotic resistance or suscep- tibility in bacterial samples. Weis conducted a study investigating this approach on key bacterial strains such as Escherichia Coli, Staphylococcus Aureus, and Klebsiella Pneumoniae, applying Ma- chine Learning methods like logistic regression, LightGBM, and deep neural networks. Despite promising results, the models have not achieved perfect accuracy, which may be due to the imbalanced database classes between bacterial resistance and susceptibility. Exploring additional strategies to improve detection in this context of class imbalance is necessary. This study analyzed the Extreme Learning Machine (ELM) machine learning technique, including two weighted ELMs proposed by Zong and the SMOTE technique, to create new synthetic samples of the minority class. Finally, by analyzing the geometric mean and accuracy performances, they achieved up to 85% accuracy and 85% in geometric mean for classification when using the weighted ELM 1 with the SMOTE technique of Oversampling.