In this context, High-Frequency Trading (HFT) helps traders hold positions for short periods of time and earn their profits by accumulating tiny gains on a large number of transactions (Huang, Huan, Xu, Zheng, & Zou, 2019). These have benefited the automation of algorithmic trades in financial instruments at very high speeds. In this domain, Machine Learning algorithms have grabbed the interest of academics and practitioners due to their ability to capture nonlinear relationships in the input data and predict price movements without relying on traditional assumptions on their statistical properties nor introducing human bias (Atsalakis, Valavanis, 2009, Hsu, Lessmann, Sung, Ma, Johnson, 2016). Apart from the use of raw prices and volumes using traditional statistical methods, the prediction of price movements can be approached using fundamental and technical indicators (Lo, Mamaysky, Wang, 2000, Tay, Cao, 2001). Market data is often characterized by the presence of noise, a high degree of uncertainty, and hidden relationships (Huang, Nakamori, & Wang, 2005). These, which are emerging as an alternative to traditional centralized currencies, have attracted significant attention in recent years due to the blockchain ecosystem and the high volatility of their exchange rates (Li, Wang, 2017, Nakano, Takahashi, Takahashi, 2018, Vidal-Tomás, Ibañez, 2018).įinancial prediction is a domain full of challenges. Convolutional LSTM neural networks outperformed all the rest significantly, while CNN neural networks were also able to provide good results specially in the Bitcoin, Ether and Litecoin cryptocurrencies.Ĭryptocurrencies are a kind of digital assets based on cryptographic protocols and technologies, such as the blockchain, that run on decentralized networks and make transactions secure and difficult to fake. The results, based on 18 technical indicators derived from the exchange rates at a one-minute resolution over one year, suggest that all series were predictable to a certain extent using the technical indicators. The experimental work compares the performance of four different network architectures -convolutional neural network, hybrid CNN-LSTM network, multilayer perceptron and radial basis function neural network- to predict whether six popular cryptocurrencies -Bitcoin, Dash, Ether, Litecoin, Monero and Ripple- will increase their value vs. This study explores the suitability of neural networks with a convolutional component as an alternative to traditional multilayer perceptrons in the domain of trend classification of cryptocurrency exchange rates using technical analysis in high frequencies.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |