Done in my senior year of high school, 2021, after completing an online introductory data science course, I eagerly applied my knowledge in pursuit of experience in the field. Although this project was done prematurely without college level understanding of research methods, I reaped tons of logistical and programming experience from my first personal project.
Inspired by the 2020 GMC “memestock” surge, the primary goal of this project is to compare the measure of herd behavior between defined classes of cryptocurrencies: meme-coins and stablecoins. The project pulls data from CoinGecko’s API and handles the data using the Python library pandas. Finally, the data is inputted into an equation of herd behavior proposed by Chang et al. Although the result was somewhat inconclusive, I can use the resulting data and methodologies to approach this topic at a different angle.
For example, in the future, I may use this data as training data in a knn model to predict whether a new cryptocurrency is a "stablecoin" or a "memecoin". I can calculate the distance of the new coin's measure of herd behavior with those in the training data and store this data in a priority queue and poll k nearest values to determine the predicted definition of the coin.
Full Project Report Here