Introducing

ARIMA-LSTM Sentiment Analysis

This project is a simple ML model that was trained on a Crypto News Dataset. The model makes an API call to a crypto news website and scrapes the latest news articles. The model then uses this data to predict the price of ETH and Impermanent Loss. There are 2 models in this project, one that uses ARIMA and another that uses LSTM.

Stack

Python

For my contributions to this project, I worked on the backend of the project. I built the backend using Python and Flask. The backend was responsible for making the API calls to the crypto news website and then sending the data to the ML model to make predictions. The backend was also responsible for sending the predictions back to the frontend.

ARIMA-LSTM Sentiment Analysis Backend in Action

Some key features of ARIMA-LSTM Sentiment Analysis include:

API call to crypto news website
Scrape latest news articles
Train ARIMA and LSTM models on the news data
Predict the price of ETH and Impermanent Loss
Display predictions on frontend

Final Thoughts

This project was a great learning experience for me and it helped me first delve into the world of machine leaarning. Even though I was mainly responsible for the backend, I am also thankful to my team who helped me understand the ML models that were built during this competition.