Machine Learning Algorithms for Signals
A little bit more about our Signal Approach
We always like to talk about our signal process, because we believe is quite unique. It was born due to our ideas, experience and AI coding knowledge.
First of all, we are programmers and we have evolved on Machine Learning algorithms, that are a field inside artificial intelligence. However, we also love trading and crypto, having a long history of using signals from different sources.
With that experience, we start collecting our own signals, storing them on a JSON file to be used (as we call it in AI) a Dataset. This help the training of our Machine Learning models, where we use a mix of Extreme Gradient Boosting (with XGBoost) and Random Forest Classifier.
We use these algorithms to validate new signals that are already filtered and use only the ones with high percentage of success. Additionally, after the signal closes, independently of the result and if it's profit or loss, we add it to the dataset to furthermore train our AI Models.
The image below is blurred on purpose:
