A Comprehensive Study on How Accurately Can A Convolutional Neural Network Detect and Classify Exoplanet Transits from Kepler/K2 Mission Light Curve Data?

Using NASA's publicly accessible Kepler and K2 datasets, this study investigates the detection and classification of exoplanet transits using convolutional neural networks (CNNs). Traditional transit detection methods often fail to handle noisy light curve data or subtle planetary signals. This study automates high-accuracy transit classification by using preprocessed time-series data to train a CNN. Using metrics like precision and ROC-AUC, the model successfully separates true planetary transits from false positives. In addition to reducing human error, this method expedites analysis and provides a scalable solution for upcoming space missions. It draws attention to how astronomy research could be revolutionised by artificial intelligence.