Abstract— Digital assets called Non-Fungible Tokens (NFTs) are used to represent things like artwork, collectibles, and in-game items and are also sold as NFTs in wide range. They are typically stored and minted in smart contracts on a blockchain and exchanged online frequently with cryptocurrencies. A digital asset's uniqueness is confirmed by the NFT. Their assessment is done in several ways, and currently, the most popular method is by using some major metrics to measure their values and demands in the future. Floor price is the most important one in all of them. The floor price represents the smallest amount of money anyone can spend to become an owner of an NFT in a specific project (a member of a project). Floor pricing are established by the people who hold NFTs inside a project. The higher the floor price, the more valuable and reliable the collection is as a whole for involving the community in a project. So, in this project the dataset is created by the price factors as numerical data and several models are used as Linear Regression, KNN, SVM, Random Forest, Confusion Metrix, in training and testing the Regression Model. Random Forest Model has given the best accuracy of 96% among all of them. A Machine Learning Web Application ‘NFT Floor Price Authenticator’ is also created to generate the best accuracy to predict the future floor price.
Keywords—Digital Asset, Non-Fungible Tokens (NFTs), Mint, Cryptocurrency, Blockchain, Floor Price, Regression, Machine Learning Web Application.
