Level: MA; BA
Responsible Person: Krzysztof Gogol (gogol@ifi.uzh.ch)
Keywords: DeFi, Machine Learning, Blockchain

This thesis involves training AI models to optimize the parameters of Automated Market Makers (AMMs), which are smart contracts facilitating trading activities on the blockchain. The objective is to minimize costs for traders while maximizing profits for liquidity providers (called "yield farmers"). The research methodology encompasses backtesting utilizing on-chain data sourced from public EVM-compatible blockchains.

20% Model Design, 70% Data Gathering, Analysis, 10% Documentation

References: