Li Lin, Ph.D.
- Visiting Assistant Professor
Education
- Ph.D. Computer Science, National University of Singapore
- B.S. Computer Science (First Class Honors), University of Leicester
Biography
For the past 10 years, Li Lin’s research has been devoted to the innovation and application of data mining, machine learning, and probabilistic reasoning methods in various domains. She developed a novel method that combines data mining and statistical methods in solving the problem of disease gene location inference and disease carrier detection. Results from rigorous experimental studies show that our method consistently produces good predictive accuracies under different conditions. In collaboration with physicians from Department of Respiratory and Critical Care Medicine, National University Hospital Singapore, she developed decision models and dynamic decision models that combine physician’s expert judgments, knowledge mined from clinical data and related literatures to perform cost-effectiveness analyses of different therapeutic options and made recommendations on the best time to implement the treatment options. The outcomes of their work were published in several prestigious conferences. As a post-doctoral fellow at the Institute of High Performance Computing (Singapore) doing dengue epidemiological research, Li collaborated and motivated local secondary school teachers and students to help with the mosquito larvae collection. The mosquito larvae collected by the schools were sent to Genome Institute of Singapore (GIS) for processing, including species identification, dengue virus detection and subtype identification, and virus RNA sequencing. The results from GIS were sent to Li for analyses and modeling to yield insights into the nature of dengue spread in Singapore and the mechanisms that could help lower the incidence and prevalence of dengue fever.
Li has worked as a lecturer and as the head of area for the business analytics program in SIM University, Singapore, for two years. Her job scope mainly focused on developing teaching materials and teaching courses in the application of data mining and machine learning techniques in market segmentation, fraud detection and credit risk assessment.
Research Interests
Li is interested in applied research in which she can innovate, enhance and customize data mining, machine learning and probabilistic reasoning methods for health care to improve patient care management and reduce cost. Her recent research focuses on the development of data-driven system of integrated models to support medical decision-making and treatment management of a disease condition known as Sepsis. The model that aims at early septic patient detection uses a dynamic Markov model to represent the set of patient states, patients move through the model according to transition probabilities that govern how likely it is to transit from one state to another. The long-term behaviour of the model provides insight into the rate of developing sepsis for different groups of patients.
