Edinburgh Research Explorer

Dr Mingjun Zhong

(Former employee or visitor)

Profile photo

Research Interests

My general research interests are computational statistics and machine learning. I am interested in devising probabilistic and statistical methods for understanding the patterns and phenomena observed in real-world data. I like to collaborate with scientists from various areas to understand their data.

Qualifications

PhD, Applied Mathematics, Dalian University of Technology, China.

Undergraduate, Applied Mathematics, Dalian University of Technology, China.

Biography

Currently I am working on the EPSRC project "IDEAL: Intelligent Domestic Energy Advice Loop" with Charles Sutton and Nigel Goddard. Working on the Machine Learning component of the project, I am interested in inferring the details of when, how and by whom energy has been used in a house by devising probabilistic models to analyse high resolution gas and electricity data.

Prior to the current project, as a research assistant, I was working on the EPSRC project “The Molecular Nose” with Mark Girolami in the University of Glasgow. My main research areas were statistical analysis of spectroscopic data, modelling and formal analysis of biochemical systems. This includes development of principled statistical methodology for inference of Raman spectra data.

Prior to this, as a post-doctoral research fellow at INRIA/IRISA in France, I was working on the project “OpenViBE”, a software platform to design, test and use Brain-Computer Interfaces. This work involved practical implementation of statistical methodologies to study the EEG signals in Brain Computer Interfaces.

I was also an associate professor at the Dalian University of Technology and Dalian Nationalities University in China. My main roles were both research and teaching.

Research outputs

  1. Sequence-to-point learning with neural networks for non-intrusive load monitoring

    Research output: Chapter in Book/Report/Conference proceedingConference contribution

  2. Pseudo-marginal Markov Chain Monte Carlo for Nonnegative Matrix Factorization

    Research output: Contribution to journalArticle

  3. Latent Bayesian melding for integrating individual and population models

    Research output: Chapter in Book/Report/Conference proceedingConference contribution

View all (19) »

ID: 16618093