|
Title: |
|
Authors:
|
|
Abstract: Nowadays, cryptocurrencies distributed over the internet are frequently used in commercial activities. However, the rapidly growing trading volume and high volatility of digital currencies make the decision-making process of investors increasingly complex. The sudden price changes of the cryptocurrency markets pose great risks to investors. Furthermore, the time and additional cost losses incurred when transferring assets between different exchanges make this complex process even more difficult. In the cryptocurrency industry, which is growing with increasing initiatives, the importance of providing solutions where investors can manage their entire portfolios on a single platform and trade quickly and easily is paramount. In response to this need, a platform has been developed that provides instant access to price fluctuations in the cryptocurrency market, where users can evaluate arbitrage opportunities between different exchanges and select the most optimal trading options. Technologies such as Java 17, Spring Boot 3.1 and WebFlux (Reactive Programming) and Microservice Architecture have been used for the backend part of the platform. The frontend of the platform is developed using React.js 18.2, Flutter 3.10, TypeScript and Redux. Various machine learning models have been tested for the asset transfer suggestions that the platform will offer to the user, and as a result, a Long-Short Term Memory (LSTM) based model has been developed that makes the most accurate predictions. According to the results of the performance tests applied on the developed platform, the transaction time has been accelerated by an average of 7.7 times compared to the manual scenario. User evaluations further confirmed that users not only find the platform powerful but also use it with an extremely easy and satisfying experience. DOI: http://dx.doi.org/10.51505/ijaemr.2025.1412 |
|
PDF Download |