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 explains how discriminant analysis can classify market observations into groups such as upward and downward price moves. It illustrates the idea with technical indicators: statistical analysis can identify variables that contribute little to…
The Grey Wolf Optimizer is presented as a population-based stochastic method for optimizing a fitness function. Candidate solutions are ranked as alpha, beta, and delta leaders, while the remaining candidates update their positions using those leaders. A…
The article concludes an implementation series on the SAGDFN adaptive graph forecasting framework, focusing on its OneStepFastGConv component. The framework constructs a sparse graph by selecting significant neighbors, then applies sparse multi-head…
The article evaluates whether the CBOE Euro and Gold volatility indexes (EVZ and GVZ) add forecasting value for XAUEUR. It describes merging the indexes with daily XAUEUR OHLC data, imputing missing values with column means, and examining correlations,…
The article addresses weaknesses in randomly sampled training labels for a market classifier, including class imbalance, serial dependence among examples, and overlap between classes. It uses principal component analysis to visualize feature structure and…
The article describes the Shuffled Frog-Leaping algorithm, a population-based optimization method inspired by memetic search and particle swarm optimization. Candidate solutions, represented as frogs, are divided among memeplexes. Within each group,…
The article explains least-squares support-vector machines for time-series regression and describes how lagged observations form predictor vectors for forecasting. LS-SVM replaces a constrained nonlinear optimization with a linear system whose kernel matrix…
The article translates fixed-width fractional differentiation from Python into an MQL5 engine and custom indicator for live market data. It explains how to generate weights recursively until they fall below a chosen threshold, reverse them to align with the…
The article introduces chronological validation for machine-learning models in trading and explains why ordinary shuffled cross-validation can leak future information into training. It outlines a workflow using historical EUR/USD data: create differenced…
The article explores whether US and Japanese 10-year government bond data can help forecast USDJPY. It compares three predictor sets: currency OHLCV alone, bond OHLCV, and a combined set. The target is a future USDJPY close, and the workflow includes joining…
This article proposes a custom MQL5 money-management class that combines a KD-tree over historical log-return and ATR coordinates with an Echo State Network (ESN) processing recent return sequences. The KD-tree is intended to identify whether current…
The document explains the Dolphin Echolocation Algorithm (DEA), a population-based optimization method inspired by how dolphins use echoes to search. Candidate solutions act as agents, and their fitness information is spread across nearby alternatives in an…
This article presents a reproducible DSP pipeline for adapting trading signals to changing market rhythms. A Hilbert-transform homodyne discriminator estimates the dominant cycle period from in-phase and quadrature components, then clamps and smooths the…
This article presents GPT as a decoder-only Transformer trained first on large unlabeled sequences and then fine-tuned with labeled examples for specific tasks. It explains causal self-attention: each token can use information from itself and earlier tokens,…
The article explains UniTraj, a framework designed to handle trajectory prediction and reconstruction by treating input as a sequence with visible and missing observations. A mask marks which agent states are known, and the model aims to reconstruct observed…
The article develops a short-period moving-average crossover idea that applies averages separately to open, high, low, and close prices, then explores whether engineered features can help identify crossover opportunities earlier. Its MQL5 data pipeline…
The article describes an Expert Advisor signal that combines price-action patterns with a GARCH(1,1) estimate of near-term volatility expansion. Seven signal modes interpret breakouts, squeeze releases, band re-entries, impulses, band walks, pullbacks, and…
The article explains Invasive Weed Optimization (IWO), a population-based stochastic method for searching continuous optimization spaces. It initializes candidate solutions randomly, evaluates their fitness, and lets each candidate produce offspring, with…
The article outlines a workflow for developing a deep-learning trading bot around MetaTrader 5. It describes connecting Python to the terminal, downloading tick data, converting ticks into price data, and using those records to train a neural network. It…
This article presents principal component analysis as a way to identify combinations of cryptocurrency assets associated with different portfolio variance levels. It introduces PCA with an image-reconstruction example, then describes gathering a basket of…
This article proposes a modular MQL5 framework for preparing financial data before machine-learning models consume it. Its pipeline design chains separate preprocessing classes for median or mode imputation, standard scaling, robust scaling, min–max scaling,…
The article presents TimeGPT, a compact transformer-style model implemented for MetaTrader 5 to forecast price changes over 24 bars. It describes storing calculations in matrices, normalizing historical price changes, and mapping them into a 256-value token…
The article explains a simplified, decoder-only transformer applied to sequential security price changes in an MQL5 signal class. It outlines positional encoding with sine waves, then describes causal self-attention: project inputs into query, key, and value…
The article develops an MQL5 class modeled on Python’s Pandas DataFrame to collect, inspect, transform, and export data for machine learning. It stores tabular data in a matrix and tracks column names separately. Because the implementation holds numeric…