Machine Learning and Optoelectronics for Automated Decision Support in Engineering Project Management

Abstract

With the development of science and technology, the emergence of machine learning methods such as neural networks has brought new ideas to cost management, and machine learning technology can greatly improve the efficiency of cost management. In addition, due to the extremely high raw bandwidth and fast signal transmission characteristics of optical systems, integrated photonics has been considered for the design of computing platforms. In this paper, we propose an optoelectronic hybrid neural network-based machine learning approach to support automated decision making in engineering project management. Specifically, based on the transmission method of optoelectronic hybrid computation, we plan the overall structure of the computational architecture, including the computing unit, the storage method and the interconnection module, and we plan the timing design of the FPGAs as well as the structural evaluation and selection of each module. The results show that the machine learning method based on optoelectronic hybrid neural network can effectively support the decision-making of project management.