Elizabeth Trenholm

Data scientist and materials engineer with over 7 years of experience in mission-critical space systems at Lockheed Martin. Expertise in analytical modeling, contamination control engineering, and advanced machine learning, developed through industry and graduate research. Proficient in Python, MATLAB, SQL, and modern ML frameworks, with proven success in building predictive models and optimizing system architectures. Recognized for effective collaboration across hardware, optical, thermal, and software teams, delivering insights that enhance mission performance and reduce risk.

University of Greenwich London, UK

Graduate Student, Department of Computer Sciences 

• Developed advanced machine learning models, algorithms, and statistical analysis tools to predict orbital paths of objects in Low Earth Orbit (LEO) and extract insights from datasets, supporting scenario analysis for identifying key improvements to readiness response.
• Designed and implemented machine learning pipelines, including Bayesian Neural Networks (BNNs), Recurrent Neural Networks (RNNs) and Genetic Algorithm architecture (GAs), to process Conjunction Data Messages (CDMs) for accurate collision risk forecasting, integrating predictive models into usable data science software products.
• Directed the development, deployment, and successful integration of predictive Deep Learning models into operational environments while collaborating within a matrixed data science team to cultivate a culture of innovation and knowledge sharing.
• Proficient in Python and ML libraries (TensorFlow, PyTorch, Scikit-learn) for building, training, and deploying predictive models leveraging Deep Learning Architecture.
• Authored and contributed to publications, internal communications, and delivering impactful data science modeling insights to the research community at large.

University of California, Berkeley Berkeley, CA

Student Researcher, Department of Astrophysics 

• Analyzed over 1,000 datasets related to high-redshift galactic clusters using advanced Python modeling techniques; findings identified key variables essential for optimizing astrophysical simulations and improving predictive accuracy.
• Partnered closely with multidisciplinary teams to contribute insights into Lyman-Alpha emission analysis, helping to identify over five critical factors influencing the efficiency of astrophysical simulations within the study parameters.
• Scheduled and facilitated regular meetings with the galactic cluster research group, ensuring alignment of analysis objectives and efficient dissemination of findings.
• Presented results of data analysis to the galactic cluster group, delivering comprehensive insights that enhanced the team’s understanding of underlying astrophysical phenomena and informed the direction of future research.