Leveraging Machine Learning for Predictive Sustainability in Business Operations: A Classification Approach to Optimize Sustainable Resource Management

Abstract

The current research focuses on the application of machine learning in assessing the possibility of predictive modeling of sustainable business practices concerning resource, waste management, and carbon footprint issues. Applying the dataset compiling 642 records from industries working on sustainable development, the study examined the performance of several classification models such as logistic regression, random forest, K-nearest neighbors, support vector classifier, and decision tree. Hence, of all the classifiers, KNN offered the highest precision, including the F1-Score of 0.5078 with Decision Tree in second place of equal precision with the F1-Score of 0.5023. Logistic Regression and Support Vector Classifier were not bad except that the Random Forest gave the lowest F1-Score of 0.4025. The paper sheds light on some important aspects like energy utilization, production of waste, carbon footprint, cost of operations, and recycling activities. Machin learning allowed classification of the business activities according to their sustainable impacts. This research provides a structure that can be used by firms to undertake more efficient practices that are less damaging to the environment. Evidently, decisions supported by statistical data will provide resource optimization, improve environmental effects, and increase compliance with corporate sustainability. The findings also benefit sustainable development as they apply the predictive modeling technique and make recommendations for operational improvement in industrial and China’s commercial sectors.