A Comprehensive Study of Cryptocurrency Trend Analysis Based on a Novel Machine Learning Technique

Cryptocurrencies are digital blockchain technology-based money that mainly got into the trend in 2015. In the present manuscript, we are finding a Machine Learning-based technique that would be best optimized to predict the prices of cryptocurrencies. We have applied Linear Regression and Bayesian models on five cryptocurrency datasets gathered from Kaggle. First, the datasets are split into testing and training data and then the model is run over all the cryptocurrency datasets. Training of every model on the training dataset is done and later it is tested and individually analyzed by checking its precision and accuracy. We then tried to obtain a comparative analysis of the models. Then we have created a tabular representation of the comparative analysis so that it can be used to determine which model is working best.