Scientists of the University of California in Santa Barbara, the University of Michigan and Stanford University developed an algorithm of machine learning ICORATING capable to find roguish ICO.
The algorithm studied the information on 2251 ICO projects. Researchers compared life expectancy and the change in price for cryptocurrency with the use of the information on ICO, including a whitepaper, a constituent team, a GitHub repository, a website and other. As a result the system learned to identify roguish ICO projects with accuracy of 83%.
Scientists mention that their development will help investors to invest money in ICO safety and to find swindlers at an early stage.
“We hope that this work will help investors to reveal swindlers and to involve more efforts to automatic assessment and the analysis of the ICO projects.”
– the description of a research states.