Articles on machine learning in trading

Restrictive Boltzmann Machines are at the basic level, a two-layer neural network that is proficient at unsupervised classification through dimensionality reduction. We take its basic principles and examine if we were to re-design and train it unorthodoxly, we could get a useful signal filter.

In our models, we often use various attention algorithms. And, probably, most often we use Transformers. Their main disadvantage is the resource requirement. In this article, we will consider a new algorithm that can help reduce computing costs without losing quality.

Convolutional Neural Networks are another machine learning algorithm that tend to specialize in decomposing multi-dimensioned data sets into key constituent parts. We look at how this is typically achieved and explore a possible application for traders in another MQL5 wizard signal class.

The article is devoted to the metaheuristic Atmosphere Clouds Model Optimization (ACMO) algorithm, which simulates the behavior of clouds to solve optimization problems. The algorithm uses the principles of cloud generation, movement and propagation, adapting to the "weather conditions" in the solution space. The article reveals how the algorithm's meteorological simulation finds optimal solutions in a complex possibility space and describes in detail the stages of ACMO operation, including "sky" preparation, cloud birth, cloud movement, and rain concentration.

We continue dwelling on the topic of social behavior of living organisms and its impact on the development of a new mathematical model - ASBO (Adaptive Social Behavior Optimization). We will dive into the two-phase evolution, test the algorithm and draw conclusions. Just as in nature a group of living organisms join their efforts to survive, ASBO uses principles of collective behavior to solve complex optimization problems.

The article considers methods of encoding initial data in hyperbolic latent space through anisotropic diffusion processes. This helps to more accurately preserve the topological characteristics of the current market situation and improves the quality of its analysis.

In this article, we will continue diving into the implementation of the ACMO (Atmospheric Cloud Model Optimization) algorithm. In particular, we will discuss two key aspects: the movement of clouds into low-pressure regions and the rain simulation, including the initialization of droplets and their distribution among clouds. We will also look at other methods that play an important role in managing the state of clouds and ensuring their interaction with the environment.

GridSearchCV and RandomizedSearchCV share a fundamental limitation in financial ML: each trial is independent, so search quality does not improve with additional compute. This article integrates Optuna — using the Tree-structured Parzen Estimator — with PurgedKFold cross-validation, HyperbandPruner early stopping, and a dual-weight convention that separates training weights from evaluation weights. The result is a five-component system: an objective function with fold-level pruning, a suggestion layer that optimizes the weighting scheme jointly with model hyperparameters, a financially-calibrated pruner, a resumable SQLite-backed orchestrator, and a converter to scikit-learn cv_results_ format. The article also establishes the boundary — drawn from Timothy Masters — between statistical objectives where directed search is beneficial and financial objectives where it is harmful.

We wrap up our look at learning rate sensitivity to the performance of Expert Advisors by primarily examining the Adaptive Learning Rates. These learning rates aim to be customized for each parameter in a layer during the training process and so we assess potential benefits vs the expected performance toll.

From ChatGPT to Gemini and many model AI tools for text, image, and video generation. Transformers have rocked the AI-world. But, are they applicable in the financial (trading) space? Let's find out.

This article shows how to represent market structure as a graph in MQL5, turning swing highs/lows into nodes with features and linking them by edges. It trains a Graph Neural Network to score potential liquidity zones, exports the model to ONNX, and runs real-time inference in an Expert Advisor. Readers learn how to build the data pipeline, integrate the model, visualize zones on the chart, and use the signals for rule-based execution.

This article lays the system architecture for a multi‑account algorithmic trading setup that operates cryptocurrency CFDs on MetaTrader 5 while respecting prop‑firm constraints. It defines three core principles—fixed dollar risk, one script per account, and centralized configuration—then details the Python–MQL5 split, the 60‑second processing loop, and JSON-based signaling. Readers get practical lot‑size computation, safety checks, and position management patterns for reliable deployment.

The article explores the Battle Royale Optimizer algorithm — a metaheuristic in which solutions compete with their nearest neighbors, accumulate “damage,” are replaced when a threshold is exceeded, and periodically shrink the search space around the current best solution. It presents both pseudocode and an MQL5 implementation of the CAOBRO class, including neighbor search, movement toward the best solution, and an adaptive delta interval. Test results on the Hilly, Forest, and Megacity functions highlight the strengths and limitations of the approach. The reader is provided with a ready-to-use foundation for experimentation and tuning key parameters such as popSize and maxDamage.