engleski

Bitcoin Price Direction Forecasting Using Neural Networks

This paper compares feedforward neural networks and logistic regression for Bitcoin direction forecasting, i.e. predicting whether the prices will go up or down in the next trading day. It uses and compares different internal and external factors from 2016 to 2021 to investigate if they can successfully predict direction of Bitcoin price movements. This paper contributes to the existing literature by defining the appropriate model for Bitcoin direction forecasting, by assessing the forecasting ability of different assets and by comparing the results through different periods including bearish, bullish and stable market conditions.

attractiveness ; Bitcon ; logistic regression ; neural networks ; macro-finance factors

nije evidentirano