About

Mengge Cao specializes in Chinese art history with a focus on the development of painting formats during the Southern Song dynasty (1127–1279). Supported by the Mary Hyde Fellowship and Princeton Institute of Regional and International Studies Fellowship, his dissertation examines how small-size paintings, measuring less than 25 cm in both width and height, gained medium specificities at the turn of the late twelfth century. He argues that these paintings facilitated interpersonal communication among emperors, imperial families, and courtiers within the imperial court. His research incorporates a quantitative analysis of nearly 1,500 entries from the 'Song Dynasty Painting Database,' a digital project supported by the Center for Digital Humanities graduate fellowship. Throughout his academic career, Cao has developed a long-standing interest in the material and conceptual processes involved in the making and perception of art objects, particularly emphasizing the significance of reprographic technologies. His master's thesis explored how printed images of plum blossoms functioned as visual analogues, promoting the principles of the 'Learning of the Way,' a Song dynasty intellectual movement that shaped the literati perspective of the world. Cao is also the co-founder and co-principal investigator of Museumverse, a graduate student-initiated startup aimed at increasing access and understanding of under-represented histories through advanced technologies. The team has collaborated with local institutions such as the Historical Society of Princeton, Drumthwacket Foundation, Morven Museum and Garden, and Princeton University Art Museum. He has conducted technical workshops on 3D scanning, augmented reality, and virtual reality hosted by the Department of Art and Archaeology. Cao received his bachelor’s and master’s degrees from McGill University in Canada and has presented his research at various conferences including the Middle Period China Humanities Conference, Keystone Digital Humanities Conference, and Harvard East Asia Society Conference.

Research topics

• Artificial Intelligence
• Machine Learning
• Computer Science
• Data science
• Psychology
• Software engineering

Selected publications

• Health system-scale language models are all-purpose prediction engines

  Nature · 2023 · 432 citations

  • Computer Science
  • Machine Learning
  • Computer Science

  to train a large language model for medical language (NYUTron) and subsequently fine-tune it across a wide range of clinical and operational predictive tasks. We evaluated our approach within our health system for five such tasks: 30-day all-cause readmission prediction, in-hospital mortality prediction, comorbidity index prediction, length of stay prediction, and insurance denial prediction. We show that NYUTron has an area under the curve (AUC) of 78.7-94.9%, with an improvement of 5.36-14.7% in the AUC compared with traditional models. We additionally demonstrate the benefits of pretraining with clinical text, the potential for increasing generalizability to different sites through fine-tuning and the full deployment of our system in a prospective, single-arm trial. These results show the potential for using clinical language models in medicine to read alongside physicians and provide guidance at the point of care.

  DOI

Frequent coauthors

• Brian D. O. Anderson

  Australian National University

  102 shared

• Anton V. Proskurnikov

  58 shared

• A. Stephen Morse

  53 shared

• Lorenzo Zino

  47 shared

• Weijia Yao

  47 shared

• Mengbin Ye

  46 shared

• Yu Kawano

  45 shared

• Héctor García de Marina

  Universidad de Granada

  44 shared

Education

• PhD, Electrical Engineering

  Yale University

  2007

• Master, Electrical Engineering

  Tsinghua University

  2002

• Bachelor, Electrical Engineering

  Tsinghua University

  1999

Awards & honors

• Mary Hyde Fellowship
• Princeton Institute of Regional and International Studies Fe…

Similar researchers at Princeton University

• Jonathan Cohenh-162
• Michael Straussh-149
• Thomas E. Shenkh-136
• Michael Levineh-131
• Romain Teyssierh-130