Optimisation of makespan of a flow shop problem using multi layer neural network
31 Mar 2020-International Journal of Computing Science and Mathematics (Inderscience Publishers)-Vol. 11, Iss: 2, pp 107-122
TL;DR: The purpose of this paper is to develop an artificial intelligence and trained a neural network model for solving the flow shop scheduling problem which gives a best jobs sequence with the objective of minimise the makespan.
Abstract: This paper presents an approach based on a multi layer neural network algorithm (MLNNA) to find a sequence of jobs for flow shop scheduling problems with the objective of minimise the makespan. The purpose of this paper is to develop an artificial intelligence and trained a neural network model for solving the flow shop scheduling problem which gives a best jobs sequence with the objective of minimise the makespan. The effectiveness of the proposed MLNNA method is compared with many problems selected from different papers. A large number of problems are solved with the present MLNNA model and it is found suitable and workable in all the cases.
Citations
More filters
TL;DR: The results shows that the workflow scheduling by the deep reinforcement learning is more effective comparing with other four single objective heuristic algorithms.
Abstract: As a convenient and economic computing model, cloud computing promotes the development of intelligence. Solving the workflow scheduling is a significant topic to promote the development of the cloud computing. In this work, an Actor-Critic architecture is utilized to solve this problem achieving the task executive time minimization under the task precedence constraint. It is similar to the list-based heuristic algorithm which includes the task prioritizing phase and task allocation phase. However, the results of the two phases interact with each other. In the task prioritizing phase, given a workflow represented as the data communication time matrix and task computation time matrix, a distribution over different task permutations by the improved Pointer network can be predicted. Then, the heuristic algorithm based on the HEFT achieves the task allocation to get the task executive time. Using negative task executive time as the reward signals, the model parameters by a policy gradient method in the first phase can be optimized. The simulation experiment is done from the task executive time, and the results shows that the workflow scheduling by the deep reinforcement learning is more effective comparing with other four single objective heuristic algorithms.
16 citations
TL;DR: This paper mainly studies the influence coefficient of talent's dynamic ability and technological innovation ability, and can judge the influence degree of human resource on enterprise according to the different influence coefficient.
Abstract: Human resource is the most important resource of an enterprise, and good talents can lead the enterprise to make continuous breakthroughs. So far, human resource has become the key to the competition of enterprises in the market. The reasonable distribution of human resource can enhance the competitiveness of enterprises. In addition, giving full play to the role of talents can promote the innovation and development of enterprises. Therefore, we should give full play to the role of talents in the development of enterprises. Through the construction of the model, we can see that human resources have a positive role in promoting the development of enterprises. The efficient use of human resources can make the development of enterprises more stable. Enterprises are composed of people, and talents are the main driving force for sustainable development of enterprises. The dynamic ability and technological innovation ability of talents are the key to enhance the innovation ability and market competitiveness of enterprises. This paper mainly studies the influence coefficient of talent's dynamic ability and technological innovation ability. We can judge the influence degree of human resource on enterprise according to the different influence coefficient.
4 citations
TL;DR: A note starting point detection framework inspired by the speech knowledge base is proposed, and the characteristics of partial tone fluctuations to detect the starting point of the note can improve the performance of the detection algorithm.
Abstract: The inherent shortcomings of the automatic translation system for spoken English are its lower accuracy and more errors. In order to improve the efficiency of automatic translation of spoken English, this paper builds an automatic translation model of spoken English based on improved machine learning algorithms based on improved machine learning algorithms. Moreover, on the basis of summarizing the signal transformation methods, feature extraction, and detection function generation methods of the existing note starting point detection algorithms, this paper proposes a note starting point detection framework inspired by the speech knowledge base, and proposes the partial note fluctuation characteristics under this framework. In addition, this paper extracts the characteristics of partial tone fluctuations to detect the starting point of the note, which can improve the performance of the detection algorithm. In addition, after the model is constructed, this paper designs experiments to analyze the performance of the model constructed in this paper and counts the research results. The research results show that the translation accuracy and translation speed of the model constructed in this paper can meet actual needs.
1 citations