چكيده به لاتين
Abstract
In recent years, blockchain technology and cryptocurrencies have become one of the hot topics in the fields of technology and finance. Blockchain, as an innovative and distributed technology, has enabled the creation of various cryptocurrencies by providing transparency and security in digital transactions. However, predicting the price of cryptocurrencies remains a major challenge due to its extreme volatility and susceptibility to various factors. In this study, we examine blockchain technology and its performance, and then apply the ANFIS-GA hybrid model to predict cryptocurrency prices. Using the adaptive neuro-fuzzy inference system (ANFIS) and genetic algorithm (GA), this model has been able to identify complex and nonlinear relationships between input data such as past prices, trading volume, and price changes and provide accurate predictions. ANFIS as a nonlinear learning tool and GA are used to optimize model parameters, including membership functions and fuzzy parameters. The results of this study show that the ANFIS-GA model has high accuracy in predicting the prices of the top ten cryptocurrencies, including Bitcoin (BTC), Ethereum (ETH), and Solana (SOL).
The analysis of the proposed investment portfolio based on the predictions of this model shows that the optimal allocation between cryptocurrencies with high growth potential and more stable currencies can reduce investment risk and increase returns. This strategy provides a balanced and efficient portfolio by using diversification and risk management. It is also suggested that by integrating input variables and using data from other markets, the prediction accuracy of the model can be increased and its results can be investigated in future research.
Keywords:
Blockchain, price prediction, cryptocurrencies, ANFIS, genetic algorithm, optimization, investment portfolio, risk management, financial data analysis.
