Discussing the article: "Neural Networks in Trading: A Multi-Agent Self-Adaptive Model (Final Part)"

[MetaQuotes]

Check out the new article: Neural Networks in Trading: A Multi-Agent Self-Adaptive Model (Final Part).

In the previous article, we introduced the multi-agent adaptive framework MASA, which combines reinforcement learning approaches and self-adaptive strategies, providing a harmonious balance between profitability and risk in turbulent market conditions. We have built the functionality of individual agents within this framework. In this article, we will continue the work we started, bringing it to its logical conclusion.

It is important to emphasize that we are assessing the effectiveness of the implemented approaches, not merely the proposed ones, as our implementation included several modifications to the original MASA framework.

The models were trained on EURUSD H1 data from 2023. All indicator parameters were set to their default values.

For initial training, we used a dataset compiled in earlier work, periodically updated throughout training to keep it aligned with the Actor's evolving policy.

After several cycles of model training and dataset updates, we obtained a policy that demonstrated profitability on both the training and testing sets.

The trained policy was tested on historical data from January 2024, with all other parameters unchanged. The results are as follows:

Author: Dmitriy Gizlyk