Cryptocurrency Trend Predictions through LSTM-Based Prediction

Cryptocurrencies are network-based digital exchange that has turned out to be a commodity with a great future scope which made it gained extreme popularity as it has stock like feature in the financial sector. It has grabbed the attention of various investors and scientists making various earlier cryptocurrencies reach the market high. These cryptocurrencies rely on blockchain technology, means that they have adapted to its various properties such as decentralization, immutability, peer-to-peer transaction and transparency which makes it almost impossible to be controlled by a Central Authority. Cryptocurrency can be considered a confusing commodity to handle for investors to acquire through mining or network user to other user exchange. Various market places have been formed for trading in various cryptocurrencies such as Bitcoin(BTC), Ethereum(ETH), XRP(XRP), Polygon(MATIC), ADA and many more. These unique characteristics of cryptocurrency cause daily price fluctuations, making it difficult to forecast their price. In this paper, taking the unique characteristics into consideration, we need a tool for prediction to help the investor learn whether the value of a cryptocurrency will go up or down, these types of prediction tools are common in stock market industry, so we can do the same for the cryptocurrency sector. Long short-term Memory (LSTM) is a complicated neural network based on ANN architecture. This paper studies about crypto currency close-price prediction of various popular cryptocurrencies using the LSTM mode