Emotions have a significant effect on human cognition processes and affect many facets of life. Interest in automatically identifying emotions has increased with the development of machine learning and artificial intelligence. Because human emotions are complex and involve physiological changes, precisely identifying them is still a challenge. Using cardiac signals from inexpensive sensors and a publicly available dataset, this study intended to develop an autonomous emotion classification algorithm that is independent of subjects. The proposed work used machine learning to process cardiac signals and classify nine different emotional states. Four machine learning models were used, demonstrating how crucial it is to take physiological variability into account for the best subject-independent outcomes. The solution's broad application potential spans from medicine to consumer product evaluation, justifying its significance.