Scientists create a crypto portfolio management AI trained with on-chain data

The system is structured into 5 primary units that work together to process info and handle structured portfolios. These modules include an information feed unit, data refinement unit, portfolio representative unit, live trading unit and agent upgrading unit. Source: Huang, Tanaka, “A Scalable Reinforcement Learning-based System Using On-Chain Data for Cryptocurrency Portfolio Management”Once established, the scientists evaluated CryptoRLPM by assigning it 3 portfolios.