Abstract

The purpose of this study is to present the application of statistical learning methods for predictive analytics within engineering management, based on enhancing decision-making and achieving operational efficiency. By using four core algorithms that were applied in the study-Neural Networks, Support Vector Machines (SVM), Linear Regression, and Decision Trees-the performance evaluation with respect to predicting major key engineering outcomes is established as follows: Results reveal that Neural Networks had an accuracy of 94.5%, showing robustness to manage complex data patterns. The second best result was SVM, with an accuracy of 91.2%, but it excelled in nonlinear data and had a greater demand for computation. Linear Regression and Decision Trees scored 78.4% and 82.3%, respectively, indicating that they were good for simplicity and interpretability but not capable with complex data. The results of the comparative analysis show that Neural Networks and SVM are good for predictive applications in engineering management. This study recommends further research into hybrid models that achieve the best balance between accuracy and interpretability to support improvements in resource allocation and productivity. This research adds value by shedding light on how statistical learning can be practically applied to engineering practice as a foundation for data-driven strategic decision-making.