Deep Learning Approach for Supply-chain Network Analysis

2016 Fiscal Year Final Research Report

• Project/Area Number: 26330344
• Research Category: Grant-in-Aid for Scientific Research (C)
• Allocation Type: Multi-year Fund
• Section: 一般
• Research Field: Web informatics, Service informatics
• Research Institution: The University of Tokyo
• Principal Investigator: MORI Junichiro 東京大学, 政策ビジョン研究センター, 准教授 (30508924)
• Project Period (FY): 2014-04-01 – 2017-03-31
• Keywords: 取引ネットワーク / 機械学習 / ネットワーク分析

Outline of Final Research Achievements

Aiming at supporting business partner recommendations and designing sustainable supply-chain networks, we propose the method to analyze customer-supplier networks using a machine learning approach. In particular, we extracted latent features from the structure of large and heterogeneous customer-supplier networks using representation learning. And we obtained the learning model which generalizes the structure of the customer-supplier networks. We also developed the business partner recommendation system based on the model. Finally, our results showed the important latent features for predicting potential business partners from a customer-supplier network. Those features can be also utilized for designing resilient supply-chain networks.

Free Research Field

知能情報学