Forex Buy Sell Signal Neural Network Powered
- Indicadores
- Ayoub Sabri
- Versão: 1.10
- Ativações: 5
DQN Multi-Forex M15 is a neural network-based trading indicator for MetaTrader 5, designed to analyze market conditions across multiple major currency pairs and display BUY and SELL signals directly on the chart.
The indicator uses a Deep Q-Network (DQN) model individually trained on each currency pair with historical data, to evaluate market behavior and automatically select the most suitable strategy.
Supported Pairs
EURUSD
GBPUSD
GBPJPY
USDJPY
Model Logic
Each neural model was trained on six trading strategies
Trend Strategy : trend following based on distance to SMA20
Momentum Strategy : impulse detection with ROC and MACD
Multi-RSI Strategy : cross-analysis of RSI across multiple periods
Breakout Strategy : Bollinger Bands and Donchian Channel breakouts
Mean Reversion Strategy : mean reversion on extreme zones
NoSignal State : unfavorable market conditions
How It Works
Automatic Detection : Pair identification and ONNX model loading
Feature Extraction : Analysis of 23 technical indicators
Market Analysis : Evaluation over a 96-bar window
Strategy Selection : Optimal choice with confidence level
Signal Generation : BUY or SELL display after candle close
Technical Features
Deep Q-Network model with 512, 256, 128, 6 architecture
Four ONNX models optimized for each pair
Six strategies integrated in training 23 technical indicators analyzed simultaneously
Full threshold calibration by percentiles
BUY and SELL signals displayed on chart
No repaint after signal confirmation
Compatible with Strategy Tester in CPU mode
Real-time information panel
Session Filters
The indicator includes trading session filtering GMT+2
Asia Session : 00h - 09h Europe Session : 09h - 18h New York Session : 15h - 24h
Sessions can be enabled or disabled individually.
Recommended Settings
Symbol : EURUSD, GBPUSD, GBPJPY, USDJPY Timeframe : M15
Disclaimer
Trading in financial markets involves risk. This indicator is an analytical tool intended to support decision-making. Past performance does not guarantee future results.
