Crypto currency portfolio allocation Using Machine Learning
Pradhyumna Rao, Nishit Bhasin, Puneet Goswami, Lakshita Aggarwal · 2021 3rd International Conference on Advances in Computing, Communication Control and Networking (ICAC3N) · 2021
Objective: Unprecedented social and economic health patterns change or when a new normal arises, people prefer to seek safety and wealth preservation. Crypto is the buzzword nowadays where people worldwide are investing their money in a digital payment transaction and have ticked all the boxes when it comes to cases such as security, seemingly fast transactions, and has quickly emerged as one the world's fastest-growing financial markets. Bitcoin, the first cryptocurrency launched in 2009 showcased the proof of work of a decentralised system and what blockchain is capable of doing in a digital ecosystem. Scope: Our paper aims at suggesting currencies to future investors based on the sentiments on popular social media and predicting the overall growth of any coin to analyze its growth prospects. Method: In this paper, we have attempted a meticulous approach to predict the price movements in cryptocurrencies considering its fluctuating factors. Since no unique model can predict the market changes effectively, we focussed on creating a SETP (Sentiment Evaluation for Trend Prediction) model, which analyzes the sentiment surrounding the crypto market and the apprehensions that engulf an investor while investing. The statistical data analysis of the top 5 coins in the market has been undertaken effectively using IBM Watson services. Further enhancements have been made using the jupyter notebook to predict the market highs of the top trending coins. The sentimental evaluation has been done by extracting Twitter trends and hashtags thereby judging the sentiments of the microblogging site and paving a way to predict the exact emotion of investors all over the globe. Findings: Finally, we believe that this research has immense capability to predict the market speculation using the various technologies that are still in the rudimentary phase and has a lot of potential in invoking the interest of researchers and providing useful data/cognizance for future investors.