MQL4 and MQL5 Programming Articles

Financial markets are not perfectly balanced. Some markets are bullish, some are bearish, and some exhibit some ranging behaviors indicating uncertainty in either direction, this unbalanced information when used to train machine learning models can be misleading as the markets change frequently. In this article, we are going to discuss several ways to tackle this issue.

As already explained in the theoretical part, when working with neural networks we need to use linear regressions and derivatives. Why? The reason is that linear regression is one of the simplest formulas in existence. Essentially, linear regression is just an affine function. However, when we talk about neural networks, we are not interested in the effects of direct linear regression. We are interested in the equation that generates this line. We are not that interested in the line created. Do you know the main equation that we need to understand? If not, I recommend reading this article to understanding it.

In the previous article, we made some fixes and added tests to our replication system to ensure the best possible stability. We also started creating and using a configuration file for this system.

The article attempts to examine financial time series from the perspective of self-similar fractal structures. Since we have too many analogies that confirm the possibility of considering market quotes as self-similar fractals, this allows us to think about the forecasting horizons of such structures.

Computer vision for trading: how it works and how to develop it step by step. We create an algorithm for recognition of RGB images of price charts using the attention mechanism and a bidirectional LSTM layer. As a result, we obtain a working model for forecasting the EURUSD price with the accuracy of up to 55% in the validation section.

The article considers Boids algorithm based on unique examples of animal flocking behavior. In turn, the Boids algorithm serves as the basis for the creation of the whole class of algorithms united under the name "Swarm Intelligence".

What we create today will be difficult to understand. Therefore, in this article I will only talk about the initial stage. Please read this article carefully, it is an important prerequisite before we proceed to the next step. The purpose of this material is purely didactic as we will only study and master the presented concepts, without practical application.

In this article we will look at a very interesting and quite challenging programming concept, although it should be treated with great caution, as its misuse or misunderstanding can turn relatively simple programs into something unnecessarily complex. But when used correctly and adapted perfectly to equally suitable situations, recursion becomes an excellent ally in solving problems that would otherwise be much more laborious and time-consuming. The materials presented here are intended for educational purposes only. Under no circumstances should the application be viewed for any purpose other than to learn and master the concepts presented.

The DeMarker Oscillator and the Envelope indicator are momentum and support/resistance tools that can be paired when developing an Expert Advisor. We therefore examine on a pattern by pattern basis what could be of use and what potentially avoid. We are using, as always, a wizard assembled Expert Advisor together with the Patterns-Usage functions that are built into the Expert Signal Class.

In this article, we will explore and implement, methods for assessing model quality that utilize a single dataset as both training and validation sets.

The article discusses the non-parametric HSIC (Hilbert-Schmidt Independence Criterion) statistical test designed to identify linear and non-linear dependencies in data. Implementations of two algorithms for calculating HSIC in the MQL5 language are proposed: the exact permutation test and the gamma approximation. The method efficiency is demonstrated on synthetic data modeling a non-linear relationship between features and the target variable.

We continue our experiment that aims to examine the behavior of population optimization algorithms in the context of their ability to efficiently escape local minima when population diversity is low and reach global maxima. Research results are provided.

This article takes a fresh perspective on a hidden, geometric source of error that quietly shapes every prediction your models make. By rethinking how we measure and apply machine learning forecasts in trading, we reveal how this overlooked perspective can unlock sharper decisions, stronger returns, and a more intelligent way to work with models we thought we already understood.

Replay buffers in Reinforcement Learning are particularly important with off-policy algorithms like DQN or SAC. This then puts the spotlight on the sampling process of this memory-buffer. While default options with SAC, for instance, use random selection from this buffer, Prioritized Experience Replay buffers fine tune this by sampling from the buffer based on a TD-score. We review the importance of Reinforcement Learning, and, as always, examine just this hypothesis (not the cross-validation) in a wizard assembled Expert Advisor.

There is a powerful and pervasive force quietly corrupting the collective efforts of our community to build reliable trading strategies that employ AI in any shape or form. This article establishes that part of the problems we face, are rooted in blind adherence to "best practices". By furnishing the reader with simple real-world market-based evidence, we will reason to the reader why we must refrain from such conduct, and rather adopt domain-bound best practices if our community should stand any chance of recovering the latent potential of AI.

This article explores the powerful role of matrix factorization in algorithmic trading, specifically within MQL5 applications. From regression models to multi-target classifiers, we walk through practical examples that demonstrate how easily these techniques can be integrated using built-in MQL5 functions. Whether you're predicting price direction or modeling indicator behavior, this guide lays a strong foundation for building intelligent trading systems using matrix methods.

How can a trader manage capital? How can a trader and investor keep track of expenses, income, assets, and liabilities? I am not just going to introduce you to accounting software; I am going to show you a tool that might become your reliable financial navigator in the stormy sea of trading.

The MacroHFT framework for high-frequency cryptocurrency trading uses context-aware reinforcement learning and memory to adapt to dynamic market conditions. At the end of this article, we will test the implemented approaches on real historical data to assess their effectiveness.

This article is somewhat different from my earlier publications. In this article, we will talk about an alternative representation of time series. Piecewise linear representation of time series is a method of approximating a time series using linear functions over small intervals.

During the offline learning, we optimize the Agent's policy based on the training sample data. The resulting strategy gives the Agent confidence in its actions. However, such optimism is not always justified and can cause increased risks during the model operation. Today we will look at one of the methods to reduce these risks.

Today, we explore another component of ALGLIB, leveraging its mathematical capabilities to develop a Polynomial Regression Channel indicator. By the end of this discussion, you will gain practical insights into indicator development using the MQL5 Standard Library, along with a fully functional, mathematically driven indicator source code.

High probability Setups are well known in our trading community, but regrettably they are not well-defined. In this article, we will aim to find an empirical and algorithmic way of defining exactly what is a high probability setup, identifying and exploiting them. By using Gradient Boosting Trees, we demonstrated how the reader can improve the performance of an arbitrary trading strategy and better communicate the exact job to be done to our computer in a more meaningful and explicit manner.

MetaTrader 5 provides rich performance data but limited structural analysis. This article shows how to export results to CSV from MQL5 and build five Python visualizations that expose cross-asset parameter consistency, the lag‑versus‑noise trade-off, walk‑forward decay, drawdown depth and duration, and intraday hour‑by‑day clusters. A unified automation module runs the full pipeline on any new export to deliver repeatable diagnostics.

Finally, our Chart Trade indicator starts interacting with the EA, allowing information to be transferred interactively. Therefore, in this article, we will improve the indicator, making it functional enough to be used together with any EA. This will allow us to access the Chart Trade indicator and work with it as if it were actually connected with an EA. But we will do it in a much more interesting way than before.

In this article, we build the core of the TimeFound intelligent model step by step, adapting it to real-world time series forecasting tasks. If you are interested in the practical implementation of neural network patching algorithms in MQL5, you have come to the right place.

Continuing tries to decipher price movements... What about linguistic analysis of the "market dictionary" that we get by converting the binary price code to BIP39? In this article, we will delve into an innovative approach to exchange data analysis and consider how modern natural language processing techniques can be applied to the market language.

In this article, we will begin to address the issue of tick excess that can impact application performance when using real data. This excess often interferes with the correct timing required to construct a one-minute bar in the appropriate window.

We will build a biologically correct system of neurons for time series forecasting. The introduction of a plasma-like environment into the neural network architecture creates a kind of "collective intelligence," where each neuron influences the system's operation not only through direct connections, but also through long-range electromagnetic interactions. Let's see how the neural brain modeling system will perform in the market.

We design a simple external trade analytics pipeline for MetaTrader 5 and implement its backend in Python with Flask and SQLite. The article defines the architecture, data model, and versioned API, and shows how to configure the environment, initialize the database, and run the server locally. As a result, you get a clean base to capture closed-trade records from MetaTrader 5 and store them for later analysis.