This research investigates the possibility of a machine learning (ML) technique for predicting and improving antibiotic effectiveness. The research aimed to identify molecular structures with high antibacterial properties by training machine learning algorithms on a huge dataset of chemical substances. The results reveal that ML models using R-squared regression descriptors outperformed earlier techniques in terms of antibacterial activity prediction accuracy. These results emphasize machine learning's potential for accelerating antibacterial discovery, hinting that it may be employed as a supplementary tool in current efforts to combat antibiotic resistance and improve therapy development.