From the neural network to the blockchain
Good morning, in this last week reading about neural connections, learning the human brain and other issues I came across a related fear between blockchain, artificial intelligence and neural networks.
I ran into MAGOS is a platform that sells tokens, but it goes beyond that (I'm not advertising this criptomoneda).
"Tuned to predict future events with high accuracy."
MAGOS seeks to combine the highly accurate forecasting capability of neural networks with blockchain technology.
This says in his presentation:
"The MAGOS fund, managed by AI and supervised by our team, will be deployed on the Ethereum platform. Initially operating in the prediction markets and sports betting houses, it will evolve and expand into other fields such as Digital Asset Management, where MAGOS forecasts will be applicable and valuable. "
What is MAGOS?
MAGOS is a complex prediction model based on a system of five Neural Networks. Each network has its own purpose and its combined powers allow to predict the results of the events with high precision, surpassing the crowd. The backbone of MAGOS is a modular architecture that supports updates and development of new features.
source
The five Neural Networks implemented in MAGOS include Strategus - a risk management network; Obelus - allows to avoid events of high level of uncertainty; Ariolus - combines the output of all other networks to make accurate prediction based on multiple layers of data; Galaxia - Network of mining and classification of data with reduction of noise, to feed information to networks of superior level; And Melius - auto tune the hyper-parameters based on the adjustments in the overall performance of the model and changes in the external environment.
MAGOS Results:
The open MAGOS test began in February 2017 and lasted 4 months. We used a third-party site with open tracking statistics. The prediction of e-sports - one of the largest growth betting markets - was chosen as a test method.
The results have shown that MAGOS is able to maintain a significant advantage over the competition, gaining net positive returns after each month of testing, with a total ROI of 28%.