Key ideas of MarketPawns: Difference between revisions
No edit summary |
|||
| Line 23: | Line 23: | ||
=== Community=== | === Community=== | ||
This key idea served as the reason for creating the public resource '''[https://marketpawns.com MarketPawns]''', is to enable analysts and traders to | This key idea served as the reason for creating the public resource '''[https://marketpawns.com MarketPawns]''', is to enable analysts and traders to easily discuss market and share ideas using not only written or verbal but also visual language .The universality and clarity of [[Universal Market Geometry (UMG)|UMG]] encourages the collaborative search for new approaches to analysis and trading. | ||
Revision as of 00:13, 29 May 2023
The driving force behind MarketPawns is the intention to create an opportunity for people to combine their mind potential with the collective intelligence and power of technology to work with such a complex, multi-level object as financial markets. The effectiveness of this will be indicated by the real profit that a user can extract by implementing tools we have created. Here are the ideas that MarketPawns has adopted as key to the implementation of the plan:
Universal Analysis System
MarketPawns uses universal system for analyzing quotes that uniformly describes almost any market data according to strictly formalized rules. It's called Universal Markets Geometry (UMG).
Models Description
For each UMG model, dozens of calculated parameters are computed, which can be used to classify any model. Using software algorithms developed over a couple of years we have compiled a database of several million models and for each one of them classification data includes more than a hundred parameters.
Forecast Evaluation System
In addition to the calculated parameters, we have developed a system of parameters to describe the price development after a model appears on the chart. This stage is referred to as "model execution". The group of parameters characterizing model execution is collectively called "control parameters".
Model Classification
By comparing calculated parameters and control parameters on the historical preice data, models can be classified by types and probabilities of forecast fulfillment. In other words, by comparing calculated and control parameters in hundreds of thousands of cases, we can predict control parameters based on calculated ones before the "execution of the model" occurs.
Neural Networks
Over several years, we have tried several approaches to analyzing the statistics of model execution. Currently, the use of specially trained neural networks shows the most promise.
Community
This key idea served as the reason for creating the public resource MarketPawns, is to enable analysts and traders to easily discuss market and share ideas using not only written or verbal but also visual language .The universality and clarity of UMG encourages the collaborative search for new approaches to analysis and trading.