Toward Developing Machine-Learning-Aided Tools for the Thermomechanical Monitoring of Nuclear Reactor Components