Search for: All records

Mining industry receives an increasing attention in green energy transition technologies in the U.S. However, little information is available regarding how the U.S. mining engineering education, particularly hard rock mining, is prepared to meet with the industry needs. This study summarizes a survey with an aim to understand the industry perspectives. The survey consisted of both closed- and open-ended questions. The survey results showed that the industry is concerned with a shortage of qualified graduates from the current U.S. hard rock mining engineering education system. The qualifications of the current education system need to be improved include engineering sciences underlying mining methods, mining design experience, mining feasibility study, the connection between theory and practice, and understanding the overall mining operation. The future desired qualifications were also suggested. Notably, the most desired ones in the next five years include an ability to acquire and apply new knowledge as needed and sufficient field experience. The survey participants, regardless of the nature of their affiliated mining companies, unanimously recommended that the collaborations between the industry and academia in the U.S. should be enhanced. Based on the survey results, the study concluded with four recommendations: (1) involve more multiple stakeholders in reforming mining education programs, (2) reinforce field experience as a key part of mining engineering programs, (3) enhance a closer collaboration between academia and industry, and (4) integrate emerging technologies (e.g., artificial intelligence/virtual reality) guided by pedagogical theories into new mining engineering curriculums. 

We introduce randomized algorithms to Clifford's Geometric Algebra, generalizing randomized linear algebra to hypercomplex vector spaces. This novel approach has many implications in machine learning, including training neural networks to global optimality via convex optimization. Additionally, we consider fine-tuning large language model (LLM) embeddings as a key application area, exploring the intersection of geometric algebra and modern AI techniques. In particular, we conduct a comparative analysis of the robustness of transfer learning via embeddings, such as OpenAI GPT models and BERT, using traditional methods versus our novel approach based on convex optimization. We test our convex optimization transfer learning method across a variety of case studies, employing different embeddings (GPT-4 and BERT embeddings) and different text classification datasets (IMDb, Amazon Polarity Dataset, and GLUE) with a range of hyperparameter settings. Our results demonstrate that convex optimization and geometric algebra not only enhances the performance of LLMs but also offers a more stable and reliable method of transfer learning via embeddings. 

76Ge can ββ decay into three possible excited states of 76Se, with the emission of two or, if the neutrino is Majorana, zero neutrinos. None of these six transitions have yet been observed. The Majorana Demonstrator was designed to study ββ decay of 76Ge using a low background array of high purity germanium detectors. With 98.2 kg-y of isotopic exposure, the Demonstrator sets the strongest half-life limits to date for all six transition modes. For 2νββ to the 0+ state of 76Se, this search has begun to probe for the first time half-life values predicted using modern many-body nuclear theory techniques, setting a limit of T_1/2 > 1.5e24 y (90% CL). 

Various theories beyond the Standard Model predict new particles with masses in the sub-eV range with very weak couplings to ordinary matter. A new P-odd and T-odd interaction between polarized and unpolarized nucleons proportional to s⃗⋅r̂ is one such possibility, where r⃗=rr̂ is the spatial vector connecting the nucleons, and s⃗ is the spin of the polarized nucleon. Such an interaction involving a scalar coupling gsN at one vertex and a pseudoscalar coupling gpn at the polarized nucleon vertex can be induced by the exchange of spin-0 pseudoscalar bosons. We describe a new technique to search for interactions of this form and present the first measurements of this type. We show that future improvements to this technique can improve the laboratory upper bound on the product gsNgpn by two orders of magnitude for interaction ranges at the 100 micron scale. 

Charge conservation and the Pauli exclusion principle result from fundamental symmetries in the standard model of particle physics, and are typically taken as axiomatic. High-precision tests for small violations of these symmetries could point to new physics. Here we consider three models for violation of these processes, which would produce detectable ionization in the high-purity germanium detectors of the MAJORANA DEMONSTRATOR experiment. Using a 37.5 kg yr exposure, we report a lower limit on the electron mean lifetime, improving the previous best limit for the e->nununu decay channel by more than an order of magnitude. We also present searches for two types of violation of the Pauli exclusion principle, setting limits on the probability of an electron to be found in a symmetric quantum state. 

Abstract TheMajorana Demonstratorwas a search for neutrinoless double-beta decay (0νββ) in the⁷⁶Ge isotope. It was staged at the 4850-foot level of the Sanford Underground Research Facility (SURF) in Lead, SD. The experiment consisted of 58 germanium detectors housed in a low background shield and was calibrated once per week by deploying a²²⁸Th line source for 1 to 2 hours. The energy scale calibration determination for the detector array was automated using custom analysis tools. We describe the offline procedure for calibration of theDemonstratorgermanium detectors, including the simultaneous fitting of multiple spectral peaks, estimation of energy scale uncertainties, and the automation of the calibration procedure.