on its surge and recommend joining the traders club. However, investment based solely on human observation and intuition in trading can cause very significant losses especially for a beginner. it is in this vision that the exploitation of historical data of the cryptocurrency and the different variables that can influence its future price. In our study we will predict the variations of a coin through different models that we will establish and then improve in oder to compare them and select the one that best estimates future values. And finally we will put our model into production.
Our goal is to build an automated decision tool that suggests the operations to be carried out in cryptocurrency trading similarly to a specialist who also knows the future.
1- Data sources exploration and Data integration
2- Data preparation and exploratory data analysis
3- Building and benchmarking Forecasting models ( ARIMA, XGBOOST, LSTM )
4- Decision making model