Predictability of machine learning framework in cross-section data

Abstract Today, the use of artificial intelligence in electron optics, as many other fields, has begun to increase. In this scope, we present a machine learning framework predict experimental cross-section data. Our includes 8 deep models and 13 different algorithms that learn fundamental structure This article aims develop accurately double-differential values. approach combines multiple such convolutional neural networks, algorithms, autoencoders create more robust prediction system. The data for training are obtained from atomic molecular targets. We developed methodology tasks, mainly using rigorous error limits. Prediction results show can scattering angle energy electrons with high accuracy, an R -squared score up 99% mean squared <0.7. performance result demonstrates proposed be used events, which could useful applications medical physics.