Brookhaven National Laboratory Develops Machine Learning Model to Accurately Identify Particle Jets in Heavy Ion Collisions

·by Henderson·AI
Brookhaven National Laboratory Develops Machine Learning Model to Accurately Identify Particle Jets in Heavy Ion Collisions
Key Points
  • BROOKHAVEN NATIONAL LABORATORY has developed a machine learning model that accurately identifies particle jets.
  • The model learns directly from simulated data without requiring labels, enhancing measurement precision.
  • Jet signals are obscured by background particles, affecting measurement accuracy.
  • The team plans to apply the model to real data and expand its use to other experiments.

Physicists at Brookhaven National Laboratory have developed a machine learning model that can automatically identify particle jets from the substantial background noise generated in heavy ion collisions. This advancement will aid in improving the measurement accuracy of quark-gluon plasma, a state of matter that existed within a microsecond after the Big Bang. The Brookhaven team collaborated with researchers from Columbia University and the University of Colorado Boulder to describe this method in Scientific Reports. The laboratory stated that this is the first time a "label-free, unsupervised" generative model has been used for this task.

The model learns directly from simulated collision data rather than relying on labels provided by scientists.

Innovative Label-Free Model

The difficulty in observing jets arises from the fact that, in the Relativistic Heavy Ion Collider (RHIC), heavy ions collide at nearly the speed of light, briefly recreating a "soup" of free quarks and gluons. Jets are concentrated bursts of high-energy particles and are one of the best tools for probing this plasma. They lose energy and change structure as they traverse the plasma. However, the signals from jets are obscured by a significant background of other particles, a challenge that has long limited the accuracy of jet measurements. Brookhaven physicist and Stony Brook University assistant professor Yan Zhu Gao stated, "Improving background reduction can make measurements of interactions between quarks and gluons under extreme conditions more precise."

”

Typically, training algorithms requires labeling thousands of examples, such as tagging images as cats or dogs. However, Brookhaven computer scientist Dmitry Tolbunov pointed out that real experimental data does not provide answer keys, making this approach impractical. The new model omits labels. It showcases pure background and ideal simulated jets, deriving how to extract jet signals while removing the background. Gao likened it to separating a crowded image into two parts: jets and everything else.

The team tested the model on simulated data from sPHENIX, a detector that operated during the last three runs of RHIC. In these tests, the model accurately reconstructed the position, energy, and internal structure of jets compared to traditional background reduction methods, especially for larger jets where background fluctuations are strongest. Current results are based on simulated data.

Future Application Prospects

The next step is to apply the model to real data. sPHENIX collected nearly 300PB of data during its operation, equivalent to billions of smartphone photos. Gao stated that this will be the first use of such a model in nuclear and particle physics. The team anticipates it will be applicable to planned electron-ion colliders and energy measurements at the Large Hadron Collider in Europe. Since the model can operate on detector data with similar images, researchers believe it may also assist other experiments that require separating overlapping signals.

This project is related to the Department of Energy's Early Career Award for the Genesis Mission Phase I, led by Baruch College, with contributions from Brookhaven, Stony Brook University, Columbia University, and the University of Colorado Boulder. Funding from Brookhaven's laboratory-directed research and development and support from the Department of Energy's Office of Science have facilitated this work. Gao praised the close collaboration with Brookhaven computer scientists, stating that this is not a one-time project, and she is excited about the new opportunities for machine learning in science.

Potential of Machine Learning in Particle Physics

This advancement at Brookhaven National Laboratory demonstrates the potential of machine learning in particle physics research, particularly in handling complex data. Traditional methods often rely on large amounts of labeled data, while the new model's unsupervised learning approach effectively overcomes this limitation, providing new possibilities for the precise measurement of particle jets. As the model is further applied, it may have profound implications for research in nuclear and particle physics and drive technological advancements in related fields.

H
About the author
Henderson