📖 Comprehensive User Guide

Complete documentation for FxMath AI Evolution v2.3.0 — Last updated June 2026

1. Introduction

FxMath AI Evolution is an advanced automated trading system that uses machine learning (LightGBM) to analyze multiple timeframes and execute trades on MetaTrader 5. Unlike traditional Expert Advisors, it features online learning — models continuously improve from live market data.

Key capabilities:

• Multi-Symbol Support: Trade unlimited pairs simultaneously through independent instances.
• Multi-Timeframe Analysis: Simultaneous M5, M15, M30 analysis with signal voting.
• Online Learning: Models retrain every 10 candles using SL/TP hit-test labeling.
• SL/TP Optimization: Dynamic ATR-based stops that adapt to market volatility.

2. System Requirements

3. Installation

The application is completely standalone. All required components (Python runtime, LightGBM, pandas, numpy, scikit-learn, etc.) are bundled in the single EXE file.

1. Download FxMath_AI_Evolution.exe from your purchase confirmation page.
2. Create a dedicated folder such as C:\Trading\FxMath_AI_Evolution\.
3. Place the EXE file in this folder and double-click to run.

4. First Launch

4.1 Terminal Selection

The app automatically detects all installed MT5 terminals on your system. Select the terminal you want to use and click "Continue with Selected Terminal". Make sure your MT5 is already logged into your trading account.

4.2 License Verification

Your MT5 account number is displayed automatically. Click "Verify License" to validate your account with the FxMath server. This is a one-time check — the app only verifies that your account is licensed; there is no expiration date.

4.3 Dashboard Overview

After successful verification, the main dashboard opens with four panels:

• Sidebar: Navigation and Start/Stop trading controls.
• Main Panel: Dashboard statistics, instance management, optimizer, and settings.
• Log Panel: Real-time trading activity, retraining updates, and errors.
• Footer: Connection status and version info.

5. Creating Instances

A Trading Instance is a complete configuration for trading a specific symbol. You can create multiple instances to trade different pairs simultaneously.

Step-by-Step:

1. Click "⚙️ Instances" in the sidebar.
2. Click "➕ Add New Instance" to open the configuration dialog.
3. Configure the following:
   • Instance Name: A label (e.g., "EURUSD-M15")
   • Symbol: The trading pair (e.g., EURUSD, XAUUSD, ETHUSD)
   • Timeframes: Space-separated list (e.g., M5 M15 M30)
   • Lot Size: Fixed 0.01 or risk-based percentage
   • SL/TP: ATR-based with multipliers 1.5× and 3.0×
4. Toggle Online Learning to enable continuous model adaptation.
5. Click Save to create the instance.

6. Trading Parameters

Understand the core settings that drive your trading success:

Risk Management

• Risk %: Percentage of account balance to risk per trade (default: 1.0%).
• SL Multiplier: Stop Loss = ATR × multiplier (default: 1.5).
• TP Ratio: Take Profit = SL distance × ratio (default: 3.0).
• Fixed Lot: Override risk-based sizing with a fixed lot size.

Machine Learning

• Online Learning: Enable automatic retraining from live data.
• Retrain Interval: Number of candles between retraining sessions (default: 10).
• Training Bars: Amount of historical data used for training (default: 500).
• Prediction Horizon: Bars forward to simulate SL/TP hit-test (default: 10).
• Confidence Threshold: Minimum prediction certainty required (default: 0.6).

Position Management

• Trailing Stop: Automatically moves SL to lock profits as price moves favorably.
• Breakeven: Moves SL to entry price + pips once profit threshold is reached.

Filters

• Time Filter: Restrict trading to specific hours or trading sessions.
• News Filter: Pause trading around high-impact economic news events.

👉 For a complete reference of every parameter including the Strategy Optimizer, LightGBM hyperparameters, indicator search space, and position management details, see the ⚙️ Settings Reference page.