The global demand for Liquefied Natural Gas (LNG) has prompted a need for optimized plant operations, particularly in the liquefaction phase where variables such as flow rate, temperature, and pressure must be effectively managed. Traditional control methods struggle with the dynamic and complex nature of these variables, leading to inefficiencies in both energy consumption and product quality. This study explores the application of machine learning (ML) and optimization algorithms to maximize LNG production by suggesting optimal sets of decision variables. Using over ten years of operational data from two LNG plants, the research integrates predictive ML models, such as decision trees and random forests, with advanced optimization techniques, including genetic algorithms and a custom random-walk algorithm. The findings demonstrate significant production improvements, with the custom ML algorithm achieving an 18.9% increase in LNG flow. This data-driven approach offers a promising alternative to traditional trial-and-error methods, enabling operators to enhance operational efficiency in LNG production.