This repository contains a comprehensive set of files and models used for classifying the time range during which an airdrop token is sold off by a wallet after a specific period of time. The training samples used in the analysis were obtained from the Optimism Airdrop 2 dataset. Below is a detailed guide to the files within the repository:
-
classified_data: This file contains labeled data from the Optimism Airdrop 2 dataset and DeepDao's List 1 and 2, based on their respective sell-off dates.
-
data: This file contains unprocessed and unlabeled raw data.
-
deepdao_metrics: This file contains account information associated with the addresses from List 1 and List 2 provided by DeepDao.
-
queried_data: This file contains data obtained from specific addresses queried using the Etherscan API, including List 1, List 2, training and testing data.
-
report: This file contains a comprehensive summary of the analysis, including the findings and methodology employed.
-
on_chain_analysis.ipynb: This is a Jupyter notebook file containing code for querying specific wallet details from Ethereum addresses.
-
model_weights.zip: This file contains saved weights for the wallet classifier model.
-
Modelling.ipynb: This is a Jupyter notebook file containing code used for feature engineering and modeling.
-
DDA.ipynb: This is a Jupyter notebook file containing code for data analysis on the addresses provided in DeepDao's List.
-
Classifier.py: This is a Python script used for running the classification model. It takes in an Ethereum address and outputs the likely sell date.