Cryptocurrency Forecasting Using Optimized Support Vector Machine with Sine Cosine Metaheuristics Algorithm




Abstract:
Predicting the market behaviour is a crucial task for cryptocurrency investors. Based on the prediction, they make decisions that will either bring profit or loss. The prediction is typically involves the historical data that is used to forecast the future behaviour of the prices on the market. Prediction is based on the machine learning approach. The nature-inspired algorithms have been successfully applied in optimization of numerous machine learning models in the recent years. Swarm intelligence metaheuristics, a family of nature-inspired algorithms, have proven to be excellent optimizers not only in the machine learning domain, but in various other practical domains as well. This paper proposes one such approach, more precisely the enhanced version of the sine cosine algorithm to optimize the support vector machine, and use it to predict the cryptocurrency prices. The basic SCA was improved with a simple exploration mechanism, and then compared to other approaches executed on the same datasets. The results obtained from the performed experimental simulations indicate that the proposed method achieved better performances than other approaches included in the research.

CITATION:

IEEE format

M. Salb, A. Elsadai, M. Živković, N. Bačanin Džakula, “Cryptocurrency Forecasting Using Optimized Support Vector Machine with Sine Cosine Metaheuristics Algorithm,” in Sinteza 2021 - International Scientific Conference on Information Technology and Data Related Research, Belgrade, Singidunum University, Serbia, 2021, pp. 315-321. doi:10.15308/Sinteza-2021-315-321

APA format

Salb, M., Elsadai, A., Živković, M., Bačanin Džakula, N. (2021). Cryptocurrency Forecasting Using Optimized Support Vector Machine with Sine Cosine Metaheuristics Algorithm. Paper presented at Sinteza 2021 - International Scientific Conference on Information Technology and Data Related Research. doi:10.15308/Sinteza-2021-315-321

BibTeX format
Download

RefWorks Tagged format
Download