OPTIMIZING MANUFACTURING PROCESSES THROUGH MACHINE LEARNING: A RANDOM FOREST AND PCA APPROACH WITH CONTINUOUS MONITORING AND RESILIENT MODEL DEPLOYMENT

Abstract

Machine learning (ML) techniques are considered an essential tool for factories to optimize their operations in various dimensions such as quality control and predictive maintenance. Hence, this article presents a strong framework that combines Principal Component Analysis (PCA) for feature extraction and Random Forest for predictive modeling as a solution to problems pertaining in multiple sectors of manufacturing. The framework accommodates the argument of Site Reliability Engineering (SRE) and Chaos Engineering to fulfill a resilient model deployment course, thus offering operational adhesion in real-time production environments. Also, the approach further employs continuous monitoring and adaptive retraining for model accuracy over time. This approach also substantially increases the operational efficiency of manufacturers, minimizes the downtime, and guarantees effective decision-making through the employment of smarter, more reliable systems. In conclusion, the framework serves as powerful mnemonics for applications in machine learning in the manufacturing industry modified to able to resist dynamic stimulus and evolving data.