Blockchain-Based IoFLT Federated Learning In A Fuzzy/Gan Environment For A Smart Trading Bot

Abstract

A DAPP is performed with collaborative training, where "Federated Learning"uses each device client to work as a singular artificial intelligence model using machine learning. The purpose is to reduce the latency by using computing resources from all client devices and increase privacy since personal data does not leave the client's devices. Applying machine learning massively in decentralized trading bots using blockchain seems to be a great solution. This learning solution can be improved using a fuzzy generative adversarial network environment to help the training. In this case, the expert system has a Python bot to interact with the Binance API to place buy/sell orders for the BTC-USD pair.