Towards the Prediction of Financial Blockchain Products using Generative Adversarial Networks

  • Dr. Kalaivany S., Dr. Artheeswari S.

Abstract

In recent days, the interesting application of bitcoin Blockchain technologies made investors focused on the return and risk rate of financial products. Therefore, it is needed to determine the return rate of bit coin in earlier. This paper introduces a novel return rate prediction model for Blockchain financial products using Generative Adversarial Network (GAN) with multilayer perceptron (MLP), called GAN-MLP.The presented GAN method is used for predicting the new return rate of Blockchain financial product. The presented GAN-MLP model undergoes training in a dedicated way for the prediction of daily closing prices by providing the historical stock data. For assessing the effective outcome of the GAN-MLP method, the Ethereum (ETH) return rate is selected as the target and performance validation takes place on the time series. The simulation results are investigated interms of two evaluation parameters such as Mean Squared Error (MSE) and Mean Absolute Percentage Error (MAPE). The presented GAN-MLP model has demonstrated superior performance by attaining minimum MSE of 0.0698 and 0.0962 on the training and testing set respectively.

Published
2020-03-30
How to Cite
Dr. Kalaivany S., Dr. Artheeswari S. (2020). Towards the Prediction of Financial Blockchain Products using Generative Adversarial Networks. International Journal of Advanced Science and Technology, 29(3), 13360 -. Retrieved from http://sersc.org/journals/index.php/IJAST/article/view/31536
Section
Articles