Description
Yield Aggregation can be done with a predictive component. A strategy manager can use a forecasting model to predict the yield from a given pool in the next time epoch. This prediction in turn impacts the asset allocation of the agent.
The first step in introducing this capability is creating a training dataset for the yield prediction algorithm. This dataset should be composed of publicly available data sources, ideally onchain. This is required for the agent to be able to fetch the newest onchain data and make a prediction based on it.
The target variable should be the APY in the next time epoch (e.g. on the next day/week).
Submission
Please create a PR in this repo with a functioning data fetching and processing pipeline, ideally using python for a streamlined integration with the Giza Agent.
Description
Yield Aggregation can be done with a predictive component. A strategy manager can use a forecasting model to predict the yield from a given pool in the next time epoch. This prediction in turn impacts the asset allocation of the agent.
The first step in introducing this capability is creating a training dataset for the yield prediction algorithm. This dataset should be composed of publicly available data sources, ideally onchain. This is required for the agent to be able to fetch the newest onchain data and make a prediction based on it.
The target variable should be the APY in the next time epoch (e.g. on the next day/week).
Submission
Please create a PR in this repo with a functioning data fetching and processing pipeline, ideally using python for a streamlined integration with the Giza Agent.