Artificial Intelligence based forecasting

2-2

The goal of our system is to find correlations between events and outcomes to ultimately create optimal forecasts.

Accomplishing this, we begin processing our data through the analysis of all the data points; partly according to numerous classic financial models (100s of various published models), but also according to our proprietary algorithms and lastly via the Artificial Intellingence. Afterwards we forward both the treated and non-treated data to our AI engines, contributing to their learning of decision trees and deep-learning neural networks.

The Artificial Intelligence engines create models to calculate various trajectories (tracks) used for extrapolation of future events in a multi-dimensional and multi-variate space (up to hundred dimensions) and probability of each scenario. Our models then separate the different types of important events from the noise, thus making it possible to distinguish the relevant from the non-relevant as well as cause and effect. These models are the basis of our predictions and are used to create our forecasts. In reality it is similar to how weather forecasts are made.

All the models are self-learning, ie, they are corrected, updated and fine tuned automatically on a daily basis as fresh data arrives. Thus, the system can create forecasts of trends with a vector which has both direction and magnitude to support the buy or sell decision.