Expanded analysis of machine learning models for nuclear transient identification using TPOT

Industries around the world are becoming more and data driven. The nuclear field is no exception with several different applications being proposed. One popular area of research use machine learning in transient detection. This paper seeks to build upon a previous study which made AutoML package TPOT train traditional models classify events occurring reactor. Synthetic was once again collected using GPWR reactor simulator. Data on 12 15 initial conditions. A dataset consisting over 100,000 points compiled used 7 pre-defined dictionary preprocessing techniques. Three trained were able produce validation results 90s expanded dataset. Once trained, it possible look into where during simulation, misclassifications occurred. Using these three models, analysis done determine if could be that effective important features missing. from this positive newly scoring close original models. Finally, conclude study, high performing retrained random states see there any major variation when used.