notes.txt

"""
# Find old SOLD listings -> Gather historical data (ETH price at listing)
# Gather ETH price at listing, ETH price at SALE, Floor price at listing, NFT traits
Training Data will include:

- Floor price
- Bid/ask history?
- Final SALE price (in ETH)

"""


# Questions to answer:
- How (what method) will we calculate NFT Rarity? (Statistical, Inverse, etc.)
- What will dataset look like? (Time-series or stationary)
    + Traits
        ~ Feature vector of length `n`, where `n` is the number of traits
    + ETH price (at time of sale)
        ~ Scalar

# Overall Procedure
1a) Pick NFT project with sufficient Liquidity using Rarity.tools or other aggregator
1b) Collect & Create training curated dataset for specific NFT projects
2a) Train the model and evaluate loop
2b) Identify mispriced NFTs inside project by inferencing forward (time-series model?)
3) Buy and Flip 'em.