Research on machine learning to reduce cost and increase efficiency in factories

Since our country has proposed the goal of "reaching the peak of carbon, carbon neutral," low-carbon reduction has become an inevitable trend. The construction and improvement of a carbon cost management system have become essential challenges that enterprises must face. For enterprises, integrating environmental considerations into their development strategy can not only address this challenge but also seize opportunities to enhance competitiveness. As major participants in the market economy and bearers of social responsibility, analyzing employees' low-carbon organizational behavior plays a crucial role in successfully implementing a low-carbon transformation strategy. In order to explore the application of data mining algorithm in enterprise cost reduction and efficiency increase, this study employs Support vector machine regression. The research involves collecting and processing business dining records along with relevant climatic data encompassing average temperature, weather conditions, maximum and minimum temperatures, and wind speed. Data harmonization and onehot coding were performed on the collected data. Subsequent to data preprocessing, Pearson correlation analyses were performed on the corporate dining records and climate data. Then, a prediction model based on Support vector machine regression models, incorporating techniques such as regularization and kernel functions to amplify prediction efficacy. The experimental outcomes underscore Support vector machine regression algorithms demonstrate feasibility and reliability. This study aims to help factories rationalize dining resources and improve dining experience.