ICO project fraud automatic discrimination method based on multimode hybrid deep learning

Abstract

The invention discloses an ICO project fraud automatic discrimination method based on multi-mode hybrid deep learning. The method comprises the following steps: (1) extracting multi-mode information of an ICO project as feature input; (2) constructing a hierarchical deep learning network, and training a deep learning model based on a current ICO project as a training sample; And (3) inputting theICO project characteristics needing to be judged into the trained hierarchical deep learning model, judging that the project is a fraud project if the output is 0, or else judging that the project isa non-fraud project. According to the method and the system, the problem of automatically judging whether the ICO new project is a fraud project or not can be solved, the defect of the method for solving the problem at present is overcome, investors can be helped to judge the fraud ICO project and prevent the fraud ICO project from fund loss, and a supervision department can be assisted in automatic early warning of ICO project fraud.

Description

technical field [0001] The invention belongs to the technical field of ICO project fraud discrimination, in particular to an ICO project fraud automatic discrimination method based on multi-mode hybrid deep learning. Background technique [0002] The full name of ICO is Initial Coin Offering, that is, initial coin offering or blockchain crowdfunding, which is essentially a financing model in the blockchain industry. Cryptocurrencies (such as BTC, ETH, NEO, etc.) are attracting more and more attention. Contrary to centralized electronic currencies and centralized banking systems, most digital tokens do not require any authorization; these decentralized systems pass The blockchain works, and the blockchain is an open, growing distributed ledger. Specifically, the issuer conducts financing by issuing tokens (Token) based on blockchain technology, and the investor pays legal currency (such as US dollar, RMB, etc.) or virtual currency (such as Bitcoin, Ethereum, etc.) to the iss...

AI Technical Summary

Problems solved by technology

[0005] Specifically, the necessity of automatic identification of ICO project fraud is reflected in the following three points: first, ICO projects often involve huge amounts of money, and a single project is basically more than 100 million yuan; second, ICO projects are not bound by countries and laws. Generally, there is no need to register a business license, there is no national trust endorsement, investors need to bear their own risks, and most investors in ICO are enthusiasts or unprofessional investors; third, there is currently a lack of automatic screening for fraudulent ICO projects method

Images