The tensile test is the most applied mechanical test to obtain the mechanical properties of polymers, which can be used in polymeric materials classification. Through a tensile test is obtained the stress-strain curve, is from which mechanical properties such as the modulus of elasticity, tenacity, and resilience of the material are obtained, which can be used to identify equivalent mechanical behaviors in polymeric materials, whether for the distinguishing plastic waste for recycling or for classifying recycled plastic material according to the content of a polymer type in its composition. However, obtaining mechanical properties from the stress-strain curve involves calculations and adjustments in the intervals of the curve in which these properties are determined, turning it into a complex process without the use of specialized software. By understanding the behavior pattern of a material’s stress-strain curve, machine learning (ML) algorithms can be efficient tools to automate the classification of different types of polymeric materials. To verify the accuracy of an ML algorithm in classifying three types of polymers, tensile tests were performed on specimens made of high-density polyethylene (HDPE), polypropylene (PP), and polyvinyl chloride (PVC). The dataset obtained from the stress-strain curves was used in the training of a multilayer perceptron (MLP) neural network. With an accuracy of 0.9261 for the test set, the model obtained from the MLP neural network was able to classify the polymers based on the stress-strain curve data, thus indicating the possibility of using an ML algorithm to automate the classification of polymeric materials based on tensile test data.