Discussing the article: "Neural Networks in Trading: Superpoint Transformer (SPFormer)"

New comment
 

Check out the new article: Neural Networks in Trading: Superpoint Transformer (SPFormer).

In this article, we introduce a method for segmenting 3D objects based on Superpoint Transformer (SPFormer), which eliminates the need for intermediate data aggregation. This speeds up the segmentation process and improves the performance of the model.

The training algorithm was inherited from previous publications, along with the supporting programs for training and evaluation.

The trained Actor policy was tested in the MetaTrader 5 Strategy Tester, using real historical data for January 2024, with all other parameters unchanged. The test results are presented below. 

During the testing period, the model made 54 trades, 26 of which were closed with a profit. This accounted for 48% of all operations. The average profitable trade is 2 times higher than the similar metric for unprofitable operations. This allowed the model to make a profit during the testing period.


Author: Dmitriy Gizlyk

 

Mate this is very interesting but very advanced for me!

Thanks for sharing, learning step by step. 

New comment