Nomis Math Model

Powered by AI

The Nomis Scorte algorithm is a powerful combination of statistical methods and machine learning techniques, including neural networks. Its purpose is to analyze data and identify dependencies between multiple parameters.

The algorithm follows a systematic approach that begins with onchain data extraction and then enriches it using third-party services. Expert judgment is also incorporated into the process.

To process time series data, the algorithm leverages various neural network architectures such as RNNs, LSTMs, TCNs, and Transformers. Both expert knowledge and neural network models are utilized for scoring, with a primary objective of enhancing accuracy through parameter tuning.

Flexible System

Nomis analyzes over 50 blockchains and 30 parameters, including, but not limited to: wallet balance, txs volume, wallet age, etc. Also, the Math Model behind Nomis Protocol could be flexibly adjusted to the needs and priorities of a certain ecosystem or project.

It means that weights of parameters counted by Nomis Protocol could change from project to project. Thus a unique architecture could be provided.

Sum up

Nomis has its own scoring algorithm—a determined set of weighted parameters ruled by AI and tuned in a way to best describe a particular wallet audience. The calculated score reflects how closely the wallet matches the ideal look-alike audience out of 100 points.