nique_372/AiDataGenByLeo

Fork 0

Niquel Mendoza b887110313 Actualizar README.md

2026-02-08 19:03:22 +00:00

4.8 KiB

Raw Permalink Blame History

AiDataGenByLeo

Build AI-powered trading systems with automatic feature generation, DSL-based configuration, and 85+ built-in features.

License Notice

The source code is made visible solely for transparency, learning, and auditing purposes.

Visible Code: Available for review
Proprietary License: All rights reserved
No Modifications: Code cannot be altered
No Redistribution: Cannot be shared or republished
Commercial Use: Restricted (see LICENSE)

Read Full License | Commercial Licensing

What's Included

Framework and Tools

DSL Parser: Define trading features using intuitive domain-specific language
Feature Generator: Automatic extraction of technical indicators, ICT concepts, news data, and more
Data Collector: Generic framework for collecting training data
Python Training Script: Train AI models using a generic script for classifiers with CatBoost, KFold, and SelectBestFeatures
Feature Factory: Over 85 features available to configure your ML dataset

EasySb - ICT EA with AI

Production-Ready EA: Complete trading system with AI integration, ONNX model included, and SetFile trained on XAUUSD M5

Main Features

DSL-Based Feature Definition

Define features declaratively without writing repetitive code:

@ New vector
{
  [Rsi_Valor][](Period=14|Applied=0)
  [Vidya_Tendencia][](CmoPeriod=9|EmaPeriod=12)
  [News_HasCpiUsaToday][]()
}

Automatic Data Generation

// Initialize feature generator
m_generator.Init("MyStrategy", features_dsl);

// Generate feature vector on each bar
m_generator.ObtenerDataEnVector(vector, curr_time);

Python Training Pipeline

# Load generated data
df = pd.read_csv('data.csv')

# Train classifier
model = train_classifier(df, selected_features)

# Export to ONNX
export_model(model, 'model.onnx')

Real-Time AI Predictions

// In your EA
if(m_predictor.CheckOpenOrder(trade_config)) {
   // AI approved - execute trade
   OpenPosition(type, entry, sl, tp);
}

Repository Structure

AiDataGenByLeo/
├── Examples/      # Library usage in an ICT EA with AI
├── Extra/         # Library to measure correlations (with integrated plot)
├── GenericData/   # Classifier, DSL Parser, Feature Factory
└── Py/            # Python script for model training

License

Full details: LICENSE.md

By downloading or using this repository, you accept the license.

Quick Start

Requirements

See the requirements.txt file (requirements for MQL5 and Python).

Installation

Clone the Git repository into shared projects via CMD
Contact me on MQL5 (user: nique_372) to be added as a collaborator with your MQL5 nickname (read-only access), which will make the repository automatically appear in your Shared Projects folder
Or fork the repository

Train Your Own Model

# 1. Generate training data
# Run EA in training mode
# Data will be saved to: TerminalPath/Files/Common/EasySbAi/data.csv

# 2. Train model
# Execute the Python script
# 3 files will be generated in TerminalPath/Files/Common/EasySbAi/
# Copy the ONNX model and scaler file to:
# AiDataGenByLeo/Examples/EAs/TTrades/EasySb/res/

# 3. Run the EA
# Execute Ea.mq5 with the training parameter set to false

Documentation

Coming soon

Commercial Licensing

Need to use this software beyond the permitted scope?

Commercial licenses available for:

Source code modification rights
Development of derivative products
Distribution to clients or end users
Integration as main component in commercial products
White-label implementations
Enterprise deployments

Contact for licensing:

Platform: MQL5 Community
User: nique_372
Profile: https://www.mql5.com/es/users/nique_372

Disclaimer

Trading involves substantial risk of loss.

This software is a technical tool, not financial advice
Past performance does not guarantee future results
You are solely responsible for your trading decisions
Always test thoroughly before deploying with real capital
Use appropriate risk management in all operations

The authors assume no liability for trading losses, system failures, or any damages arising from the use of this software.

Contact and Support

Platform: MQL5 Community
Profile: https://www.mql5.com/es/users/nique_372/news
Issues: For bug reports or questions

4.8 KiB Raw Permalink Blame History