Research On Evaluation Model Of Cryptocurrency Competitiveness

Cryptocurrency has developed rapidly in recent years,but the quality of various types of cryptocurrency are not uniform,which brings a severe challenge to users in making reasonable investments and effective supervision of regulatory agencies.Constructing an evaluation model of competitiveness of cryptocurrency currency will help users to make efficient investments,and the regulatory agencies will carry out targeted supervision,which is of great significance in safeguarding the national security and financial market order.At present,there are few studies on the level of competitiveness of cryptocurrency all over the world,and the relevant data are not sufficient to reasonably evaluate the value and potential of cryptocurrency.What's more,due to the large differences between cryptocurrency and traditional currency and short lifetime of cryptocurrency,cryptocurrency indicators are very limited.This paper constructs a set of evaluation indicators that are not affected by the differences in the characteristics of cryptocurrency and we also aim to make sure the easiness of collecting data of these indicators.On this basis,a variety of competitiveness evaluation models are constructed using factor analysis,traditional machine learning methods,and social media big data methods.The competitiveness rating of 26 cryptocurrencies was predicted.Factor analysis method lacks supervised learning data,and the effect is poor compared to traditional machine learning methods.In traditional machine learning methods,the support vector machine model and neural network model are most effective.The neural network model combined with Twitter social media big data has improved the accuracy compared to traditional machine learning models.Finally,this paper discusses multiple combinations based on neural networks and social media big data,and finally achieves 76.9%competitiveness prediction accuracy on 26 encrypted digital currencies.The experimental results show that the experimental model proposed in this paper can accurately and efficiently predict the competitiveness of encrypted digital currency,and it has great application prospects in the future.