Grade: 100 Applied several ML models to attempt to classify cryptocurrencies and predict prices for the top currencies by market cap. Note that prediction accuracy was very low, so please do not attempt to use such predictions in a real market.
The training data was OHLC price data from yahoo! finance for the time period 5/2021 - 5/2022
Hypothesis: Crypto markets crashed in May 2022, after a flourishing year of bullish activity. The machine learning models were unable to forecast prices effectively because the training data had never encountered a massive drop in prices. In the past, cryptos fell but kept on making ATHs over and over again, skewing the predictive models. It would be a good idea to retest once the market stabilizes.