Quantitative Analysis on Company Rankings Using Neural Network

本研究的目的是进行实证分析，神经网络是否可以很好地作为一种方法来分析排名的决策结构。美国的穆迪投资者服务公司和标准普尔公司以及日本的排名研究机构和评级投资信息中心，为了阐明排名的决策结构，使用神经网络分析了排名决策结构的决定因素，并对制造业进行了分析。这些分析提供了高度准确的信息，发现比现有分析方法获得的信息准确得多。然而，这些信息在金融业务中的特点还没有得到明确。我希望今后继续研究使用这种方法的金融机构。

The purpose of this research was to conduct positive analysis on whether a neural network can function well enough as a method of analyzing the decision structure of ranking. A neural network was used to analyze the determinants of the decision structure of ranking in Moody's Investors Service and Standard & Poor's Corp. of the United States or the ranking research institute and the rating and investment information center of Japan, aiming to elucidate the decision structure of ranking, and was also used to analyze the manufacturing industry. These analyses provided highly accurate information that was found to be much more accurate than the information obtained with the existing analysis method. However, the characteristics of this information in the finance business have not been clarified yet. I would like to continue to research financial institutions using this method in the future.