Articles
Neural Networks in Practice: When to Use an Artificial Neuron and How It Works in MQL5 for MetaTrader 5
We implemented a neuron in MQL5 that was trained using stochastic gradient descent and compared its loss function to that of linear regression. Using code and graphs, we demonstrate how normalization, the choice of learning rate, and the structure of the problem affect convergence. This article
Neural Networks in Practice: Stochastic Gradient Descent for MetaTrader 5
This article uses practical examples to explain how to calculate and apply gradients of the weights and bias (intercept) in a linear neuron implemented in MQL5, and also discusses a stochastic variant of gradient descent. We analyze stopping criteria, iteration limits, and the effects of training on
Neural Networks in Practice: Gradient Descent for MetaTrader 5
In this article, I will try to explain as simply and clearly as possible one of the most controversial issues that arises when discussing neural networks: how to find the best possible point, or the minimum of the loss function. I'll show you the difference between linear regression and gradient
Neural Networks in Practice: An Example of Implementing an XOR Gate for MetaTrader 5
In this article, I will try to show you, dear reader, that things are not always what they seem. We often tend to think that everything is arranged in a certain way, even though we may end up believing something that isn't necessarily true. Neural networks are undoubtedly one of the most interesting
Replay and Market Simulation: The Grand Finale for MetaTrader 5
I know that many of you may have thought I would publish a few more articles to explain other aspects of this system. The missing elements are easy to implement. Even so, developing them will let you see how prepared you really are
Market Simulation: Unity Is Strength (III) for MetaTrader 5
In this article, I will present our system for simulating market operations. Although everything is practically finished, there are still a few things to implement and a few changes to make. However, I have to admit that, after everything we've already developed, I'm tired of still being stuck on
Market Replay: Unity Is Strength (II) for MetaTrader 5
Until now, the application being developed as part of this series of articles has focused exclusively on simulating the graphical part. However, to obtain a more complete system in which we can test the Expert Advisor within the replay/simulation service, we also need to simulate the trading server
Market Replay: Unity Is Strength (I) for MetaTrader 5
We're entering the home stretch. The development of the replay/simulation system is nearly complete. Of course, we still have a few things left to finish, but compared to everything we've already done, completing what's left won't be difficult. However, it is essential to fully absorb and understand
Market Simulation: Position View (XIX) for MetaTrader 5
One of the issues that bothered me the most was that the `C_ElementsTrade` class contains code for accessing positions. Don't take this as a mistake, because it really isn't one. However, this increases the risk of errors in some of the tasks we will need to handle later. All work on implementing
Market Simulation: Position View (XVIII) for MetaTrader 5
In this article, I have shown—in the clearest possible way—how to modify and improve code capable of handling specific tasks while making as few changes as possible to the existing code. We will add a volume display and, at the same time, ensure that users or traders cannot effectively remove