Discussing the article: "Neural Networks in Trading: Two-Dimensional Connection Space Models (Final Part)"
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
Check out the new article: Neural Networks in Trading: Two-Dimensional Connection Space Models (Final Part).
After completing the implementation of our own interpretation of the approaches proposed by the authors of the Chimera framework, we proceed to the final stage of our work - training and testing the models on real historical data.
To train the models, we used a training dataset collected during the training of the previously discussed models. This training dataset was built using historical data of the EURUSD currency pair for the entire year 2024 on the M1 timeframe. All indicator parameters were set to their default values. A detailed description of the training dataset preparation process can be found at this link.
Testing of the trained models was carried out in the MetaTrader 5 Strategy Tester on historical data from January 2025, while keeping the other training parameters unchanged. The testing results are presented below.
According to the test results, the model was able to generate a profit. More than 70% of the trades were closed with a profit. The profit factor was recorded at 1.53.
However, several points should be noted. The models were tested on the M1 timeframe. At the same time, the model executed only 27 trades, which is quite low for high-frequency trading on the minimal timeframe. Moreover, the model opened only short positions, which also raises questions.
Author: Dmitriy Gizlyk