A New Algorithm for Optimizing the Parameters of a High-Performance Neural Network

This study explores the potential of using artificial neural networks to identify the rare process $$pp \to tHbq$$ at the Large Hadron Collider, aiming to improve the separation of the signal and background. A neural network-based mathematical tool was developed to improve signal cleaning. This approach was validated using Monte Carlo simulation of signal and background events. We find that neural networks suggest a promising technique for increasing the significance of the signal, facilitating the detection of the process $$pp \to tHbq$$