Cryptocurrencies have recently become a widely recognised way for many individuals to invest their money. Cryptocurrency values have increased dramatically during the last several years. As a result, I tried numerous ways to find an efficient and accurate model to predict the price using machine learning algorithms. For the course project, I analysed four cryptocurrencies and attempted to predict the price of bitcoin. The CSV time series data is uploaded to GitHub. The time-series data in CSV format is posted to GitHub. The GitHub repository is then cloned into a Jupyter notebook under the Google cloud platform's AI platform. The time-series graph and the prediction graph are then saved as png files and uploaded to Github and to a bucket on the Google cloud platform that has been created separately.
- Analyzed and predicted the effect of the pandemic on the price of specific cryptocurrencies like BTC, ETH and LTC.
- Implemented AR, ARIMA and ES algorithms in Python using statsmodels library. Additionally, explored Facebook's prophet library.
- Deployed the model on Google Cloud Platform’s AI Platform and stored the result in a GCP bucket linked to Github