ALFA Transformer AI: Free MT5 Expert Advisor

Overview

ALFA Transformer AI operates by communicating via HTTP with a local Python daemon that processes machine learning inference. The system includes an advanced Kelly Criterion position sizing engine, an adaptive scalping strategy, and time-based trading filters. It requires external training of models using provided Python scripts. Users can choose between regression-only or combined trading modes with ATR-based risk management and volatility-adjusted stop losses.

LSTM and Transformer neural network integration

External Python daemon for real-time AI inference

Kelly Criterion dynamic position sizing

Adaptive scalping strategy with pip thresholds

ATR-based volatility adjusted risk controls

Details

Author

syngyn on GitHub

Type

Trading robot

Platform

MetaTrader 5 (MQL5)

Strategy

AI / machine learning, Scalping

Markets

Forex

Timeframes

M1-M5, M15-H1

Risk profile

No grid, martingale or hedging logic was found in its code or README. That says nothing about how it performs: backtest it first.

EA files

Topics

adaptiveaieaexpertadvisorlearninglstm-neural-networksmachine-learningmetatrader-5metatrader5mql5

Source

github.com/syngyn/LSTM-Future-5-bar-prediction-MQL5

Input parameters

Read from SAM (1).mq5. Defaults are the author's; change them in the EA's Inputs tab.

Name	Type	Default	Description
`DaemonHost`	`string`	`127.0.0.1`	Daemon server host
`DaemonPort`	`int`	`8888`	Daemon server port
`HttpTimeoutMs`	`int`	`15000`	HTTP request timeout (ms)
`MaxRetryAttempts`	`int`	`3`	Max retry attempts for failed requests
`RetryDelayMs`	`int`	`1000`	Delay between retries (ms)
`EnableHttpLogging`	`bool`	`true`	Enable detailed HTTP logging
`TestDaemonOnStart`	`bool`	`false`	Test daemon communication on startup
`EnableKellyCriterion`	`bool`	`true`	Enable Kelly Criterion position sizing
`MaxKellyFraction`	`double`	`0.25`	Maximum Kelly fraction (25% recommended)
`MinKellyFraction`	`double`	`0.01`	Minimum Kelly fraction (1% minimum)
`KellyLookbackTrades`	`int`	`30`	Number of recent trades for Kelly calculation
`KellyMultiplier`	`double`	`0.5`	Kelly fraction multiplier (0.5 = Half-Kelly)

Show all 40 inputs

Name	Type	Default	Description
`UseConfidenceScaling`	`bool`	`true`	Scale Kelly by prediction confidence
`BaseRiskWhenNoHistory`	`double`	`2.0`	Risk % when insufficient trade history
`EnableKellySmoothing`	`bool`	`true`	Smooth Kelly changes to prevent whipsawing
`KellySmoothingFactor`	`double`	`0.3`	Smoothing factor (0.1-0.5 recommended)
`EnableTradingHours`	`bool`	`false`	Enable/disable trading hours filter
`TradingStartHour`	`int`	`8`	Start trading hour (0-23)
`TradingEndHour`	`int`	`18`	End trading hour (0-23)
`TradingMonday`	`bool`	`true`	Trade on Monday
`TradingTuesday`	`bool`	`true`	Trade on Tuesday
`TradingWednesday`	`bool`	`true`	Trade on Wednesday
`TradingThursday`	`bool`	`true`	Trade on Thursday
`TradingFriday`	`bool`	`true`	Trade on Friday
`TradingSaturday`	`bool`	`false`	Trade on Saturday
`TradingSunday`	`bool`	`true`	Trade on Sunday
`AvoidNewsHours`	`bool`	`false`	Avoid major news hours
`NewsAvoidanceHours`	`string`	`14:30-15:30`	Hours to avoid (format: HH:MM-HH:MM)
`ClosePositionsOutsideHours`	`bool`	`false`	Close positions when outside trading hours
`EnableScalpingStrategy`	`bool`	`true`	Enable/disable scalping
`BaseScalpingPips`	`double`	`10.0`	Base pip threshold for scalping
`ScalpingTimeoutBars`	`int`	`3`	Bars to hold scalping position
`ScalpingRiskPercent`	`double`	`1`	Risk % for scalping (separate from main)
`AdaptiveScalpingThreshold`	`bool`	`true`	Make pip threshold adaptive
`MinScalpingAccuracy`	`double`	`0`	Minimum accuracy required for scalping
`ScalpingOverridesMain`	`bool`	`false`	If true, scalping replaces main strategy
`AdaptiveLearningMode`	`ENUM_ADAPTIVE_MODE`	`ADAPTIVE_CONSERVATIVE`
`EnableDynamicConfidence`	`bool`	`true`
`EnableAdaptivePositionSizing`	`bool`	`true`
`EnableStepWeighting`	`bool`	`true`

How to install it

1Download the sourceUse "Download source (.zip)" above, or clone syngyn/LSTM-Future-5-bar-prediction-MQL5 from GitHub, and unzip it.

2Open the data folderIn MetaTrader choose File, then Open Data Folder, and go to MQL5/Experts.

3Copy the filesCopy the .mq5 file into Experts.

4Compile in MetaEditorOpen the file in MetaEditor and press Compile. Fix any missing includes the compiler reports.

5Test before you tradeRun it in the Strategy Tester, then attach it to a demo chart and allow algorithmic trading.

Third-party code. ForexR did not write, test or endorse this EA, and the summary above was generated automatically from its README and code. ForexR lists only repositories with readable source code and never hosts files. Read the code, backtest it and use a demo account first; automated trading can lose money quickly.

Similar expert advisors

Gold EA MT5 With OpenAI

by RobertN1D

This Expert Advisor for MetaTrader 5 is specifically designed for XAUUSD trading on the M1 timeframe. It utilizes a hybrid strategy combining classic technical indicators like RSI, Moving Averages, and ATR with an OpenAI API integration for signal validation. The robot features an on-chart dashboard for telemetry and i

MT5 Trading robot ScalpingTrend followingAI / machine learning

Updated 5 months ago None stated GitHub

GGTH AI Modules

by syngyn

This collection of MetaTrader 5 expert advisors integrates external machine learning models via a local HTTP daemon for trade execution. The core system, GGTH-SR, utilizes sophisticated position sizing based on the Kelly Criterion with confidence-based scaling and smoothing. It includes an adaptive scalping module with

MT5 Trading robot AI / machine learningScalping

Updated 1 year ago None stated GitHub

Scalping Robot ML

by Red-misst

This expert advisor for MetaTrader 5 performs scalping by identifying price levels based on a configurable lookback period. It incorporates a machine learning signal filter, requiring a separate Python environment to classify high-probability trade setups using historical market data. The robot manages entries within s

MT5 Trading robot ScalpingAI / machine learning

Updated 1 year ago None stated GitHub

Quantum AI Trader Pro

by ExMsg

This machine learning expert advisor for MetaTrader 5 utilizes ensemble and deep learning models to analyze market patterns. It offers a versatile multi-strategy approach, allowing users to toggle between trend-following, scalping, and grid-hedging behaviors. Risk controls are comprehensive, featuring fixed percentage

MT5 Trading robot AI / machine learningMulti-strategyScalping High risk

Updated 11 months ago MIT GitHub

Gold Scalper AI Bridge

by almightymoon

This MetaTrader 5 expert advisor acts as a bridge between the trading terminal and an external Python-based AI server. It extracts market data and technical indicators—including EMA, RSI, ATR, and ADX across multiple timeframes—to request trade signals via ZeroMQ, TCP, or file-based communication. The EA manages trade

MT5 Trading robot AI / machine learningScalpingIndicator crossover Archived

Updated 4 months ago None stated GitHub

Conquest EA

by MunchonGithub

Conquest EA is a self-adjusting expert advisor for MetaTrader 5 that utilizes Smart Money Concepts such as Fair Value Gaps and Break of Structure for entries. It features a built-in learning module that evaluates its own trade outcomes to adjust strategy weighting and confidence in real-time. Risk management includes a

MT5 Trading robot SMC / ICT / price actionAI / machine learningTrend following

Updated 4 weeks ago Apache-2.0 GitHub