This project aims to predict the next day's and next week's trading volume of cryptocurrencies using a Supervised Machine Learning model. The model utilizes historical trading data and features such as date-related information to make predictions.
Project components:
- README.md: This file, providing project documentation and instructions.
- ARBtradingVolumePrediction.ipynb: Python script containing data analysis, model training, and visualization.
Requirements:
- Flipside account (API)
- Coingecko API
- python, pip, jupyter notebook (any other working environment)
Instruction:
- create an SQL query to fetch historical trading data on Flipside
- create and activate a virtualenv
Working notebook Data Source: This analysis uses data from the Flipside Crypto API, fetched from:
"https://flipsidecrypto.xyz/api/v1/queries/50f51f57-95e8-4267-a404-3b2f0acfe93a/data/latest"
Step-by-Step Analysis:
- import Libraries & Load Data:
- import necessary libraries, fetch trader data from the API, and create a pandas DataFrame.
Data Processing Download Query Results:
- Use the Flipside API to download the query results.
- Load the query results into a Pandas DataFrame.
Define Features and Target Variable:
- Identify the features (independent variables) and the target variable (dependent variable) from the DataFrame.
Split Dataset:
- Split the dataset into training and testing sets. A common split ratio is 80% for training and 20% for testing.
- Ensure that the split is random and maintains the distribution of the data.
Model Training and Evaluation Training the Model:
- Train the model using the training dataset.
Evaluating the Model:
- Evaluate the model's performance using the testing dataset.
- Use appropriate evaluation metrics: Mean Squared Error (MSE), Mean Absolute Error (MAE).
Visualization:
- Plot the predicted vs. actual trading volumes to visually assess the model's accuracy.
Testing the Model:
- Use the trained model to make predictions on new, unseen data.
- Validate the predictions to ensure the model's robustness.
Running the Analysis:
- clone Repository: Clone this repository.
- execute Script: Run the script.