Using Artificial Neural Networks to Distinguish Neutrons and Gamma Rays in Liquid Scintillator
Author Information
Author(s): Yun Eungyu, Choi Ji Young, Kim Sang Yong, Joo Kyung Kwang
Primary Institution: Center for Precision Neutrino Research, Department of Physics, Chonnam National University, Gwangju, Republic of Korea
Hypothesis
Can artificial neural networks improve the efficiency of pulse shape discrimination for neutron and gamma signal separation in the nonlinear region of photomultiplier tubes?
Conclusion
The study found that the artificial neural network method can effectively distinguish between neutron and gamma signals in a nonlinear region, achieving a classification accuracy of 99.5%.
Supporting Evidence
- The artificial neural network achieved a classification accuracy of 99.5%, outperforming traditional methods.
- Neutron signals had broader tails than gamma signals, which helped in their differentiation.
- The study utilized both 10-inch and 2-inch photomultiplier tubes to validate the method's efficacy.
Takeaway
This study shows that we can use smart computer programs to tell the difference between two types of tiny particles, neutrons and gamma rays, even when the signals they make are a bit messy.
Methodology
The study used artificial neural networks to analyze signals from photomultiplier tubes detecting neutrons and gamma rays in a liquid scintillator.
Limitations
The study noted that experimental uncertainty could affect the performance of the artificial neural network.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website