Welcome to the NinjaTraderIndicator repository developed by ML Trades, an innovative trading organization. Our indicator is designed to perform various calculations for efficient data capture and enable accurate predictions for trading decisions.
The flagship component of our indicator, APMasterIndicator, plays a pivotal role in capturing data for training AI models. Versions up to v39 are specifically designed for data collection, providing a solid foundation for AI-driven analysis.
AP5, AP7, AP14, and AP21 are essential slave indicators, working in conjunction with APMaster. These indicators perform calculations on Renko Charts with intervals of 5, 7, 14, and 21 respectively. Each slave indicator captures data in real time, storing it meticulously in a database on a bar-by-bar basis.
APMaster excels in real-time data analysis, leveraging the captured data from the database. It performs intricate statistical calculations and writes the results into a comprehensive CSV file. This file, containing an extensive 106 columns of valuable data, serves as the training set for our supervised learning models, enabling accurate predictions.
We continuously enhance APMaster to optimize its prediction capabilities. In versions APMasterv40, APMasterv41, and APMasterv42, significant updates have been implemented. Instead of storing data in a CSV file, these versions directly send data items to our dedicated AI server using Web Requests.
Our AI server, powered by a Python script built with Flask, effortlessly handles web request acceptance and provides real-time prediction responses to the Ninja Trader Script. We have employed advanced machine learning models, including RandomForest and LightGBM, for accurate trade entry point prediction. Furthermore, the state-of-the-art RESNET101 model excels in trade type prediction.
For your convenience, we have consolidated the necessary resources in a Google Drive folder. You can find the following items there:
- Data: The captured trading data used for training our AI models.
- Models: The trained machine learning models for prediction.
- Results: Comprehensive results and analysis based on our indicator's performance.
- Tutorials: In-depth tutorials to help you navigate and leverage our indicator effectively.
Feel free to explore the repository, utilize the power indicator, and make informed trading decisions backed by cutting-edge AI technology.