Time Series Methodology in STORJ Token Prediction

The STORJ token has gained a lot of attention from consumers, businesses, investors and open security software communities due to its decentralized nature and return payment ideology. Things become even more complicated if the data has high volatility. While there has been significant research done to analyze the network topology of these kinds of cryptocurrencies, limited work has been performed to analyze the efficacy of different time series models on the overall price prediction. In our work, an extensive study on linear models as well as hybrid models and temporal convolutional neural networks have been done based on different uncertainty quantification measures to show how different models perform in capturing the underlying volatility and overall price prediction.