Apedul

We think everybody who has been buying nft has faced the same problem. Which are not knowing which one is the right one. So we propose "Apedul - new solution of finding your best suit nft".
With Apedul not only you use it to find your best suit nft. You can also play it like an akinator to kind of guess your nft by their corresponding traits.

We are inspired by the Akinator guessing app. To give the user a list of questions, implement our machine-learning algorithm. We call it a binary decision tree. First, we pull all nfts in the collection (bored ape) from IPFS, then we put it in a Python data frame and use traits as columns. Token IDs are rows. The value of each column is 1 or 0, depending on whether they contain the trait or not (1 means they have the trait; 0 for not containing this trait). Next, we find the column with a ratio value near 1:1 the most, then use it as the trait to ask the user. After the user answers, we cut down the data frame by dropping the row that does not match the given trait value from the user. With this, we have cut the data frame down by half since the trait we ask has a ratio value near 1:1 the most (meaning almost half of the data). So we continued this step until we left the last row to be the best-suited nft for our users.