Classification of Company Credit Rating Using Artificial Neural Network with Data Factorization
Journal Title: International Journal of the Computer, the Internet and Management - Year 2018, Vol 26, Issue 3
Abstract
The objective of this study is to classify company credit rating applying an Artificial Neural Network (ANN) model incorporating with a Principle Component Analysis (PCA) technique for the purpose of extracting common factors from a large panel of input variables in order to overcome the multicollinearity problem in the financial figures. The rating reports are collected from a provided website of TRIS Rating (Thailand) Co., Ltd in September, 2018. The data set consists of 183 companies classified into 6 different levels of rating (AAA, AA, A, BBB, BB, and lower than B respectively). To construct the ANN model, 60% of the sample will be used as a training set and the remaining will be assigned to play a role as a testing set. With a pre-adjustment process, the Augmented Dickey Fuller test will be applied to each time series of the 26 selected financial figures. Then, a panel of standardized financial figures will be extracted. With 26 nodes of input layers and 53 nodes in 3 hidden layers, 80.91% of the company’s credit rating was correctly classified with the training set. In addition, the overall accuracy of the proposed ANN model was improved by 3.28% when they were applied with the testing set. With the empirical findings of this study, we can infer that the ANN model with data factorization would effectively provide constructive results in classifying the company credit rating.
Authors and Affiliations
Krisada Khruachalee
Keywords
Related Articles
- EP ID EP601001
- DOI -
- Views 195
- Downloads 0
The Knowledge Measurement Model Provided to the ASEAN Economic Community
The purposes of this research aim to analyze the proficiency knowledge preparation step readiness for the member country of ASEAN Economic Community and to determine the questions quality for the knowledge level preparat...
Spectrum Refarming in Thailand
Spectrum is a limited resource, but it can be reusable depending on spectrum management strategies. It is an underlying mechanism for communication and economic and social development. In order to utilize the spectrum at...
System Architecture for Green University Resource Planning on Cloud Computing
This research aimed to study and develop the design of the system architecture for green university resource planning (GURP) on cloud computing. The research was divided into two phases: 1) the design of the system archi...
A Classification Model for Predicting Standard Levels of OTOP’s Wood Handicraft Products by Using the K-Nearest Neighbor
The aim of this research is to develop a classification model for predicting standard levels of OTOP’s wood handicraft products by using the K-Nearest Neighbor (K-NN). To develop candidates of classification models, we u...
Environment Factors Related to Career Development for Disabled Students
Environment factors related to career development for disabled students consists of both external and internal environments. For external environment, it includes laws for the disabled; the government has to design and p...