Leveragıng On Contınual Machıne Learnıng For Real Tıme Tradıng In Fınancıal Markets

Authors : Emmanuel Ndidi Osegi, Vincent Ike Anıreh, Ifeanyi. C Emeto, Noble. A Ibiobu, Osu Joshua Orove, Uchechukwu Agi

Pages : 178-185

Doi:10.53070/bbd.1575411

View : 2 | Download : 0

Publication Date : 2024-12-25

Article Type : Review Paper

Abstract :In this research paper, a novel neural inspired approach called the Auditory Machine Intelligence (AMI) is proposed for the prediction of cryptocurrency price - the ripple (XRP/USD) trading pair. The AMI is implemented as part of an LUNO API systems app and real-time data generated and captured. Comparisons were also made with a popular method used in cryptocurrency trading called the Non-linear Auto-Regressive Neural Network (NARX-net). The experimental results considering the 40%, 60% and 80% training were reported for the both techniques. From the experimental results, it was clearly seen that on the average, the NeuroAMI will outperform the NARXnet technique at the 60% and 80% input training data levels; however, the NARXnet technique fared better on the average at the 40% input training data level.
Keywords : Blockchain, cryptocurrency price, machine learning, neural network, prediction