A new machine learning technology tested by Fermilab scientists and collaborators can spot specific particle signatures among an ocean of LHC data in the blink of an eye, much faster than standard methods. Sophisticated and swift, its performance gives a glimpse into the game-changing role machine learning will play in making future discoveries in particle physics as data sets get bigger and more complex.
In his doctoral thesis, Todd details a method for data analysis in a way that minimizes a source of bias in some particle physics experiments. By analyzing information from two distant detectors simultaneously rather than sequentially, he incorporated the lack of precision knowledge in both detectors. A University of Cincinnati graduate, Todd used data from Fermilab’s MINOS and MINOS+ experiments, and his analysis can be applied in other neutrino research as well.
A Michigan-based foundry was recently given an award for its casting of a prototype of a Fermilab particle accelerator component. Their method, which uses a 3-D-printed casting mold, allows for an economical approach to creating accelerator parts and could lead to significant cost savings in component fabrication.
A Fermilab group has found a way to simulate, using a quantum computer, a class of particles that had resisted typical computing methods. Their novel approach opens doors to an area previously closed off to quantum simulation in areas beyond particle physics, thanks to cross-disciplinary inspiration.
For the first time, scientists have demonstrated that low-energy neutrinos can be thoroughly identified with a liquid-argon particle detector. The results, obtained with the ArgoNeuT experiment, are promising for experiments that use liquid argon to catch neutrinos, including the upcoming Deep Underground Neutrino Experiment.