Abstract
This is a visual representation of the abstract.
The intelligent manufacturing industry is gradually replacing traditional manufacturing, and investing in intelligent manufacturing projects faces many risks. To address the insufficient investment risk analysis in manufacturing projects, an intelligent investment risk assessment method is proposed. The novelty lies in the combination of expert methods and big data mining techniques to construct project risk indicators, which improves the effectiveness of risk assessment. Meanwhile, a risk prediction model combining convolutional networks and long short-term models is introduced to analyze project investment risks and improve the accuracy of risk supervision. In the model performance test, when the sliding window was 4, the ROC area of the research model was 0.9366, indicating that the overall performance of the research model was better. The comparison of root mean square errors in model training showed that the model trained on K1 and K2 data had root mean square errors of 0.008 and 0.017, respectively, which were superior to other models. When comparing data from different partitions, this research model effectively analyzed time series data, with an overall prediction accuracy of 97.65% compared with other models. In different levels of risk prediction, the research model had the highest overall prediction accuracy, with an accuracy of 94.32%, which was better than other models. Finally, in the comprehensive risk prediction comparison, 16 experiments were conducted. The average accuracy of the research model was 94.95%, which was better than the other three models. Meanwhile, the highest and lowest predicted values of the research model were 96.48% and 93.45%, respectively, which were superior to other models. The research content can provide valuable references for enterprise investment decision-making and risk management.
Get full access to this article
View all access options for this article.