Key Data for Enhancing Deep Learning in Wide-Field-View Atmospheric Cherenkov Telescopes

This dataset encompasses a comprehensive collection meticulously curated to advance Deep Learning (DL) methodologies within the Wide-Field-View (WFV) Atmospheric Cherenkov Telescopes (ACTs), specifically tailored for the HADAR experiment set to be established in Tibet, China. It integrates the expansive coverage characteristic of traditional Extensive Air Shower (EAS) array detectors with the heightened sensitivity of focused Cherenkov detectors, targeting the observation of transient astrophysical phenomena such as gamma-ray bursts and gravitational wave counterparts.The dataset is constructed upon a foundation of innovative integration of relevant physical theories with the latest AI technologies, aiming to significantly boost the HADAR experiment's sensitivity. It includes:1. Simulated Shower Events: Detailed simulations of atmospheric showers, generated by cosmic gamma rays and other cosmic particles, providing a rich basis for training DL models to discern between various types of incident particles.2. Calibration and Environmental Data: Information on the telescope's response to known light sources and environmental conditions, essential for calibrating the DL algorithms to accurately process raw sensor data.3. Observational Data: Real observational data collected by the prototype, including targeted observations and survey data, crucial for validating the DL models' effectiveness in real-world conditions.4. Processed Outcomes: Results from applying the trained DL models to the test dataset, demonstrating background identification accuracy, energy reconstruction fidelity, and angular resolution capabilities.This dataset not only showcases a breakthrough in applying AI to astrophysical detection, with significant improvements in background identification accuracy, relative energy reconstruction error, and angular resolution but also establishes the HADAR experiment as a pioneering venture in the domain of WFV ACTs.