Few weeks ago, we wrote about Forecasting Uniswap V3 Volume Using Facebook Prophet Data Science Framework and this week let’s look how to predict Uniswap V3’s TVL using Google BigQuery ML and Flipside Crypto Data Warehouse.
Before we go any further, let’s make sure we understand few terms
- Uniswap : Uniswap is the hottest Defi application on Ethereum blockchain. It is basically a decentralized finance protocol that is used to exchange cryptocurrencies. It facilitates automated transactions between cryptocurrency tokens on the Ethereum blockchain using smart contracts. No humans or order books are involved!
- TVL: TVL (total value lock) is the one of the most popular metrics used for measuring performance of a DeFi(Decentralized Finance) applications. TVL represents the total amount of money, mostly measure in US Dollars, in a DeFi applications smart contracts.
- Flipside Crypto: One of the finest web3 organizations providing on chain data to everyone with a mission to power community enabled analytics.
- Google BigQuery: Google’s big data platform that empowers SQL Developers to build machine learning models easily.
To predict Uniswap’s TVL for next months, we are going to extract daily TVL values from Flipside Crypto and use those values to train a time series regression model on BigQuery ML framework. Once the model is trained, we will use it predict values for next 3 months.