Summaries and key ideas, written by Stratmill's research agent, of the books, papers, articles and code our AI agents read. Each page links to its original.
5,018 documents
The article presents a MetaTrader 5 trading system based on PatchTST, a transformer-style model that divides time series into patches to represent local patterns while modeling relationships across a longer history. Its described input uses normalized price…
The article describes an MQL5 trading model that combines neural-network training with Markov market-state labels, a 256-dimensional state-space model, and mechanisms named resonance, interference, and decoherence. It describes inputs spanning prices,…
This article presents a practical framework for deciding when to move MQL5 calculations from the CPU to a GPU through OpenCL. It identifies batch workloads with many similar, independent operations—such as parameter optimization, hypothesis testing, pattern…
The article presents an unsupervised Isolation Forest monitor for execution quality, built natively in MQL5 and attached to trade transaction events. Each confirmed fill is represented by features for slippage, request-to-fill latency, spread at execution,…
The article presents financial forecasting as a search for structure in noisy market data and contrasts the efficient market view with the possibility of short-term predictability. It proposes using neural networks to identify useful inputs and learn…
The article explains AutoBots, an encoder-decoder transformer originally designed to predict multiple agents’ future trajectories, and adapts its ideas to forecasting currency price movements. The encoder alternates attention across time and across elements…
This article explores ten signal patterns that combine a moving average, used as a trend indicator, with the Stochastic Oscillator, used for momentum. It frames the patterns as a supervised learning task: neural networks are trained in Python on labelled…
The article describes a stock trend prediction model that combines short-term market features with longer-term temporal structure. A convolutional neural network extracts local relationships among market variables, while piecewise linear regression segments…
This article examines how to combine technical strategies using weighted votes and genetic optimization. In an earlier experiment, the optimizer selected two strategies whose returns were highly correlated, undermining the intended diversification. The…
The article explains DeepAR, a recurrent neural network approach that learns jointly across related time series and produces probabilistic forecasts. It describes how the model uses past observations autoregressively, can incorporate categorical and…
This article introduces statistical arbitrage through correlation, cointegration, and the Pearson coefficient, then demonstrates a screening workflow using historical prices. Its Python example compares instrument pairs, retains those above a correlation…
This article outlines the software structure of an automated optimizer for MetaTrader strategies, focusing on its graphical interface and project organization. The application is split between a DLL and a separate executable, with an MVVM design that…
This article explains GSM++, a hybrid approach for representing graph-structured data through graph tokenization, local encoding, and global sequence encoding. It compares node or edge tokenization, which retains element-level detail at higher computational…
This article describes an encoder component for a Structured Component Neural Network, an approach that separates time series into long-term, seasonal, short-term, coupled, and residual parts. Its focus is extracting the coupled component: relationships…
This article develops a reinforcement learning extension to an Expert Advisor that uses ADX and CCI patterns. It contrasts a more continuous feature representation, where indicator conditions are encoded separately, with a compact representation that…
This article’s final part assembles the MASA framework from a market observer, a reinforcement learning agent, and a controller agent. The observer forecasts market sequences, the RL agent proposes portfolio actions, and the controller adjusts those actions…
This article explains how to connect the Fast Artificial Neural Network library to MQL4 through a wrapper that hides C pointers and variable-length arguments from the trading platform. It outlines creating feedforward networks, training them incrementally,…
The article introduces supervised backpropagation and shows how matrix and vector operations can express neural-network forward passes and weight updates in MQL5. It reviews neurons, layer weights, activation functions, bias inputs, and gradient descent,…
The article investigates a reinforcement-learning trading system that stopped placing trades during continued training. It frames this as a model learning failure and recommends diagnosing the behavior before changing the training setup. One proposed control…
The article explains dropout as a way to reduce co-adaptation among neural network features during training. At each training pass, a randomly selected subset of neurons is masked, encouraging the remaining features to contribute more independently. The…
This installment explains how to add a clustering step to an automated workflow for selecting groups of trading strategy instances. It uses K-means from scikit-learn to cluster results from the first optimization stage, then excludes same-cluster instances…
The article explains TiDE, a multilayer perceptron architecture for long-term time-series forecasting. It combines a history of the target series with historical and future covariates, encodes them into a dense representation, and decodes forecasts with a…
This article presents a Hidden Semi-Markov Model (HSMM) for detecting market regimes while estimating how much time may remain in the current state. Unlike a standard Hidden Markov Model, which implies geometrically distributed regime durations, the HSMM…
This article explains how to generate a neural network Expert Advisor with the MQL5 Wizard and Hlaiman EA Generator. It treats a chart pattern as a sequence of bars preceding a profitable price move, then uses graphical buy and sell signals to create…