Ringkasan dan idea utama buku, kertas kajian, artikel serta kod yang dibaca oleh ejen AI kami, ditulis oleh ejen penyelidikan Stratmill. Setiap halaman memautkan sumber asal.
3,012 dokumen
The article explains how MetaEditor’s MQL5 Wizard assembles an Expert Advisor from reusable modules without requiring the user to write the initial program. The workflow selects a trading-signal module, a trailing-stop method, and a money and risk management…
The article frames trend-system design around three practical challenges: recognizing a trend after confirmation delay, setting a profit objective before entry, and deciding whether a pause signals trend failure or a temporary flat period. It then presents…
The article introduces a self-adaptive moving average implemented as an adaptive finite impulse response filter using normalized least mean squares (NLMS). Instead of fixed moving-average weights, the filter predicts the current price from past inputs,…
The article extends a Python trading simulator designed to reproduce parts of the MetaTrader 5 interface. It explains how to retrieve historical ticks and bars and describes wrapping many built-in functions for symbol data, orders, positions, account…
The article explains MACD as the difference between faster and slower exponential moving averages, paired with a smoothed signal line; the histogram represents their difference. It gives the conventional default periods and walks through the calculation…
The article describes an MQL5 detector for double tops and bottoms that separates higher-timeframe market context from trade timing. A swing engine classifies trend and supplies confirmed swing points on a configurable swing timeframe. The detector checks…
This article surveys short-term forex entry and exit tactics based on price breaks, trendline crossings, support and resistance, corrections, and gaps. It discusses staging entries around a breakout and adding to positions during a pullback, while noting…
This installment in a market replay and simulation series explains how to separate position visualization from an Expert Advisor while maintaining communication between them. The EA provides the link to the trading server, while an indicator displays open…
This article outlines a pipeline for exporting custom backtest and optimization data from MetaTrader 5 to CSV, then using Python libraries to examine results beyond the platform’s standard reports. Its proposed normalized dataset includes instrument and…
This article describes extending an MQL5 optimization-results application so users can select a ranking criterion and view the corresponding balance curves together. A drop-down control chooses the criterion, and highlighting table rows adds their results to…
The article implements temporal features in MQL5 to match a Python feature pipeline for machine-learning models. It covers Fourier encodings for hour, weekday, and day of year; flags for major forex sessions and their overlap; session-conditioned rolling…
The article translates the Darvas box method into an MQL5 Expert Advisor. It defines a box from a price extreme that holds for three sessions, then requires a breakout above the box boundary with increased volume. The implementation adds filters for low…
The article describes a universal MetaTrader 4 and 5 template that runs robots for multiple symbols from one chart. It explains the motivation: simplify setup, reduce the load of many chart instances, and evaluate a combined multi-instrument equity curve.…
The article compares neural-network optimization methods for quickly retraining feed-forward models, with an emphasis on Levenberg–Marquardt (LM) for online adaptation. It explains output-layer error calculation for different loss functions and activation…
The article develops a recursive probability framework intended to represent price movement within symmetric, asymmetric, and one-sided corridors. It defines corridor width and midpoint, tracks the current distance from the start of a path, and accumulates…
The article describes TrendMap, a signal-generation framework that combines swing-based Fibonacci retracement levels with VWAP. It calculates a selected bar range’s high and low, derives retracement levels, and compares VWAP with the midpoint of that range:…
The article explains how legacy and updated MQL4 tools create and update offline charts from standard chart data. Its example uses a period-conversion script on EURUSD one-minute bars to generate custom two-, three-, or four-minute history, which is then…
This installment in a replay-system series explains how a chart-trading indicator can work with a mouse indicator and update displayed values through MetaTrader 5 chart and template behavior. It shows how user-provided values are passed into the chart-trade…
This article explains how to move indicator calculations from a separate MetaTrader 5 indicator into an expert advisor. The motivation may include packaging the strategy in one program, controlling which calculations run, or keeping implementation details…
The article describes a MetaTrader 5 expert advisor designed to manage trades across multiple currency pairs and two metals from one chart. Its entry logic combines a seven-period ADX directional movement signal with Parabolic SAR confirmation. The ADX rules…
The document explains why principal component analysis can obscure the importance of individual features when many inputs are highly correlated. PCA can capture shared variation and suppress noise, but its components distribute influence across related…
The article presents a workflow for building a transparent financial machine learning model and connecting it to MetaTrader 5. It contrasts glass-box models with black-box methods, discusses the disagreement that can arise among black-box explanation…
The article explains how to use ZUP’s graphical interface in MetaTrader to mark market waves and attach charting tools to ZigZag extreme points. Users can create and save multiple wave layouts for a symbol, adjust wave levels and display options, and apply…
The article compares an EURUSD hourly LSTM model trained on unfiltered prices with versions trained after low-pass filtering, then explores training on February data concatenated across years. It also varies the filter settings and prediction horizon across…