n_TOF Transport Code Update and RF Deconvolution

The Transport Code constitutes a software tool to process outputs from Monte-Carlo simulation in neutron physics research, specifically catering to n_ToF. This software generates diverse histograms pertinent to neutron physics investigations. An essential feature of the Transport Code is its capability to derive the Resolution Function (RF), a crucial component for accurately determining the true energy of neutrons. This is imperative due to the convolution of neutron energies by the neutron generation process and subsequent requirement for meticulous processing to extract authentic energy values. This scientific report revolves around the comprehensive advancement of the Transport Code framework. The advancements encompass a spectrum of activities, ranging from migrating the graphical user interface (GUI) from Python 2.7.5 to 3.6.8, to substantial enhancements in the underlying C++ codebase governing the optical simulation tool. Furthermore, this report elucidates an antecedent exploration involving the fusion of Machine Learning and neutron physics. Specifically, the application of autoencoders to implement transpose convolutions (deconvolutions) on data from detectors is discussed. The outcome of this endeavor is deliberated upon, alongside the progression achieved through the evolution of the model architecture, transitioning from an autoencoder approach to a multi-layer Convolutional Neural Network (CNN) design. Notably, the outcome of these codebase enhancements is the attainment of results identical to those produced by the legacy code, albeit with the advantages of contemporary robustness and maintainability. Concurrently, the Machine Learning investigation exhibits promising potential through the proposed model, pending exhaustive testing and validation, for practical implementation in real-world scenarios.