Ai in the consensus
AdminKeymaster# 4 years, 3 months ago
Question from Ivan, Ivan on Tech interview
Q Ivan: How do you use AI in your consensus and how would you explain your consensus algorithm to us? Could you use POW and POS as example, or is it completely unrelatable?
A BIllal: Our new consensus is proof of involvement and integrity. This means that to elect and select the bot, which will validate the block on the blockchain, we will use matrix based on their involvement and integrity. We will use thermodynamic parameters for this.
The calculation of involvement of each bot in the network will be based on the parameters entropy and enthalpy. In neurochain entropy means the level of the interaction of each bot with his peers and enthalpy is about the quality of exchanges between the bots. These parameters rely on the involvement of the bots. The third parameter is integrity. Integrity is about how the bot is interacting in the network. For integrity, we are using machine learning algorithms, such as bayesian network. Example of this is: You are a bot and you want to interact with another bot. For this interaction it will calculate the probability of this interaction, depending on your past interactions. When you do ‘normal’ interactions, there won’t be any issues and the probability will be high. The problem comes when you do strange/out of the order interactions, then the probability will decrease and your score of integrity will be low.
You must be logged in to reply to this topic.