This notebok contains 2 parts to the project.
Formulation of 3 datasets:
- Historical Data: This tracks bitcoin indices using the investpy library
- Tweets collection: This is just a collection of all the tweets analyzed
- Rate Checker: This tracks the sentiment in comparison with the bitcoin rates.
Data Analysis and Building a ML model: As this was only scoped for about 10 days, the size of the dataset is small so the ML model performs with a below par accuracy. But given enough data it can scope out to one with higher accuracy. The ML model predicts a binary variable "Rise" that is 1 if the price rises and 0 if it drops.