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.
3,012 documents
The article describes three complementary methods for reducing the chance that a strategy’s backtest reflects historical noise: Validation-within-Validation (V-in-V), Combinatorially Purged Cross-Validation (CPCV), and Combinatorially Symmetric…
The article explains how a binomial distribution models the count of successes across a fixed number of independent trials with a constant success probability. It maps wins and losses to successes and failures, then describes an MQL5 visualization tool that…
This article presents HedgeTerminal, a MetaTrader 5 panel and library that represents a trader’s net obligations as separately manageable long and short positions. MetaTrader 5 normally aggregates exposure into a net position, while MetaTrader 4 supports…
The article explains money management as control over the loss a trade can incur, rather than as a formula that scales lots only with account balance. It argues that strategy performance depends on win rate, average win, and average loss, and that losses…
The article develops a reusable MQL5 framework for finding and displaying price patterns in timeseries data. It proposes an abstract base pattern class with specialized pattern classes, plus management objects that let a timeseries search for multiple…
The article proposes replacing fixed RSI thresholds with levels based on the indicator’s observed range in the market being traded. Its motivation is that conventional overbought and oversold readings may be absent for extended periods, while simply changing…
The article explains the design of an MQL4 library for common matrix operations. It stores each matrix as a row-major, one-dimensional array and keeps row and column counts separately, then groups functions for resizing, row and column manipulation,…
This article extends a Strategy Builder based on Merrill patterns with additional testing and reporting options. The prior version summarized results in points; the update lets users measure profit in points or account currency, set an initial deposit, and…
This article replaces an augmented Lagrangian optimizer in an MQL5 volatility-model library with Sequential Least Squares Programming (SLSQP). The motivation is a mismatch between parameter estimates from the MQL5 implementation and Python’s ARCH module,…
The article adapts the deterministic Dendritic Cell Algorithm, originally developed for anomaly detection, to continuous optimization. It maps candidate solutions to antigens, fitness to danger and safe signals, and agents’ maturation cycles to decisions…
The article explains how an MQL5 Expert Advisor can obtain indicator handles and read indicator-buffer values. It covers two routes: creating a handle with IndicatorCreate() and using a technical-indicator function. IndicatorCreate() accepts indicator type…
The article examines whether random number generator quality affects stochastic population optimizers. It distinguishes pseudorandom generators from true random sources, reviews common generator families, and describes a 64-bit Mersenne Twister…
The article presents the Electromagnetism-like algorithm, a population-based metaheuristic for unconstrained continuous optimization. It models candidate solutions as charged particles: objective-function scores determine attraction or repulsion, and the…
The article explains SARSA, an on-policy reinforcement-learning method, and describes using it as an independent signal model in an MQL5 Expert Advisor. Its Q-value update uses the action actually selected in the next state under the agent’s current policy,…
This article presents a reusable execution layer for MQL5 Expert Advisors. It separates transient submission failures, such as requotes or temporary connection loss, from permanent rejections such as invalid stops or insufficient margin. A retry executor…
This installment explains the first part of an MQL5 order-management class that receives chart-trading events and sends market buy or sell requests to a trading server. It describes decoding event data, selecting the contract, reading its bid or ask and…
This article describes risk controls and volatility estimation for emulated option strategies. For long option structures, it adds an early exit when the underlying reaches a selected rebalancing level; the trader can choose the level and stop trading for…
Chaos Game Optimization (CGO) is a population-based method for searching multidimensional spaces. Agents begin at random positions within defined bounds. Each agent samples a subgroup, calculates its average position, and generates a candidate using one of…
The article outlines a pipeline for characterizing multifractality in EURUSD five-minute returns. Starting from partition-function slopes calculated in an earlier installment, it extracts the scaling function tau(q), estimates the Hurst exponent using…
This article describes changes needed to add a SimpleCandles strategy to a modular, multi-currency Expert Advisor project and its automated optimization workflow. It emphasizes keeping reusable library code independent from project-specific logic, then…
This article explains how to coordinate components in an MQL5 Expert Advisor without shared global variables. It proposes a publish-subscribe event bus: components publish typed events, and registered listeners receive the events they need through a central…
The article describes an ensemble for a moving average channel Expert Advisor. The channel uses moving averages of high and low prices to generate breakout signals, while a Ridge Regression model uses the same technical indicators. Trades are allowed only…
The article explains batch normalization as a way to normalize neuron activations during training, aiming to reduce changes in the distributions passed between network layers. It describes calculating a batch mean and variance, standardizing values, then…
The article describes a proposed MetaTrader 5 strategy that combines a deep learning model’s next-close predictions with sentiment scores derived from news headlines and descriptions. In the backtest, a position is taken only when the direction implied by…