A Study on Productivity Analysis Based on Big Data Analytics and ARIMA Modeling

This study explores the application of machine learning algorithms in multidimensional data analysis, especially the performance of ARIMA model in modeling and forecasting of new productivity-related data, using entropy weight method and TOPSIS comprehensive evaluation method. First, a scientific quantitative input system is constructed for the diversified data indicators, covering the fields of green technology and digital economy. Secondly, focusing on the time series prediction using the ARIMA model, the high fit and prediction accuracy of the model in complex data are fully verified, highlighting its practicality in the field of machine learning. Finally, the contribution of each data was assessed by combining statistical analysis and comprehensive evaluation methods, which verified the important role of the ARIMA model in improving the efficiency of data analysis and prediction accuracy. The findings indicate that the algorithm not only enhances the predictive accuracy significantly, but also provides reliable support for decision optimization in multidimensional data analysis.