Discussing the article: "Neural Networks in Trading: Generalized 3D Referring Expression Segmentation"

[MetaQuotes]

Check out the new article: Neural Networks in Trading: Generalized 3D Referring Expression Segmentation.

While analyzing the market situation, we divide it into separate segments, identifying key trends. However, traditional analysis methods often focus on one aspect and thus limit the proper perception. In this article, we will learn about a method that enables the selection of multiple objects to ensure a more comprehensive and multi-layered understanding of the situation.

During the training process, we applied an algorithm previously validated in our earlier studies.

The trained Actor policy was tested in the MetaTrader 5 Strategy Tester using historical data from January 2024. All other parameters remained unchanged. The test results are presented below.

During the test period, the model executed 22 trades, exactly half of which were closed in profit. Notably, the average profit per winning trade was more than twice the average loss per losing trade. The largest profitable trade exceeded the largest loss by a factor of four. As a result, the model achieved a profit factor of 2.63. However, the small number of trades and the short testing period do not allow us to draw any definitive conclusions about the long-term effectiveness of the method. Before using the model in a live environment, it should be trained on a longer historical dataset and subjected to comprehensive testing. 

Author: Dmitriy Gizlyk