Researchers from the University of Tsukuba in Japan have developed an AI-powered cryptocurrency portfolio management system called CryptoRLPM. The system uses reinforcement learning to incorporate on-chain data into its model and is made up of five primary units.
The scientists tested CryptoRLPM by assigning it three portfolios and found that it demonstrated significant improvements over baseline performance. The system showed at least an 83.14% improvement in accumulated rate of return, at least a 0.5603% improvement in daily rate of return, and at least a 2.1767 improvement in Sortino ratio compared to the baseline Bitcoin.
All Comments