Abstract:
Motivation-Due to the availability of Industry 4.0 technology, the application of Big Data analytics to automated systems and pure optimization problems is possible. Problem statement-The majority of studies in optimization fields mainly focused on optimizing the objectives despite the long solving time. Although solutions from ANN can be used immediately as final results, the overall accuracy after training state is still limited, so in the context of this research, one more step will be conducted by setting outputs from ANN as initial solutions to solve those complex optimization problems easier and faster. Purpose-This study aims to find the solutions of each optimization problem mentioned in introduction within a shorter solving time by applying ANN to get the initial solution for them. Design/Methodology-The combination between Machine Learning technique for Optimization will be applied to those complicated problems, namely, air cargo packing problem, vehicle routing problem (VRP), and dynamic path planning. Result-The successfulness of applying ANN to predict the solutions and set them as initial ones gives the shorter time when solving the optimization models. Originality-The contribution of this research is to provide a new approach and encourage a new research direction in solving traditional operation research and can be considered using in automated system with some machine learning techniques
Thammasat University. Thammasat University Library