Discussing the article: "Neural Networks in Trading: Hyperbolic Latent Diffusion Model (Final Part)"

New comment
 

Check out the new article: Neural Networks in Trading: Hyperbolic Latent Diffusion Model (Final Part).

The use of anisotropic diffusion processes for encoding the initial data in a hyperbolic latent space, as proposed in the HypDIff framework, assists in preserving the topological features of the current market situation and improves the quality of its analysis. In the previous article, we started implementing the proposed approaches using MQL5. Today we will continue the work we started and will bring it to its logical conclusion.

Training is conducted using real historical data for the entire year of 2023 on the EURUSD instrument with the H1 timeframe. All indicator parameters were set to their default values.

The training process is iterative and includes regular updates to the training dataset.

To verify the effectiveness of the trained policy, we use historical data for the first quarter of 2024. The test results are presented below.

As the data shows, the model successfully generated a profit during the testing period. A total of 23 trades were executed over the course of three months, which is a relatively small number. Over 56% of the trades were closed profitably. Both the maximum and average profit per trade being approximately twice as large as their loss counterparts.

Author: Dmitriy Gizlyk

 
MetaQuotes:

Check out the new article: Neural Networks in Trading: Hyperbolic Latent Diffusion Model (Final Part).

Author: Dmitriy Gizlyk

It is technologically impressive but the practical outcome is rather modest.
New comment