Analyzing Consumer Attitudes Towards Sustainable Products using SVM and Autoencoders on E-Commerce Platforms

On a global and national level, there is an effort to encourage more sustainable food consumption. Despite the field's rapid expansion in the past decade, studies examining consumer perspectives and purchasing behaviour are few and far between. One must be knowledgeable about customer beliefs, morality, and preferences in order to adjust to evolving lives and interests. In order to evaluate customer behaviour, a SVM ensemble model incorporates WLPPD for every base classifier. In order to test the model, we use a number of Sustainable Lifestyle Rating datasets. By incorporating novel characteristics and filtering out less-than-ideal hidden-layer outputs during training, HDSSAs improve learning. Compared to existing techniques, the HDSSA model performs better on the Sustainable Lifestyle Rating Dataset. A binary classification F1macro score of 0.91 demonstrates its capacity to detect eco-friendly purchasing habits. Findings demonstrate that state-of-the-art machine learning can increase comprehension of environmentally friendly eating habits. By outlining a thorough framework for studies on consumer sustainable behaviour, this paper improves both theoretical and applied research. Sustainable food marketing, policymaking, and uses of multi-class categorization should be the focus of future research.