Researchers apply innovative mathematical technique to help combat obesity
Researchers from Katz School and UMass Dartmouth have unveiled a groundbreaking method that employs an advanced mathematical technique to deepen understanding of the factors influencing weight loss. Detailed in their white paper, “A Choquet-Integral Based Approach to Identify Weight Loss Component Subsets,” their research was showcased at the IEEE/ACM International Conference on Connected Health: Applications, Systems, and Engineering Technologies (CHASE) in June.
Traditionally used in theoretical mathematics and economics, the Choquet Integral is now making strides in the biomedical field. This technique provides a sophisticated method for targeting and optimising various metrics and factors affecting weight loss, showing promise in the practical application to health data.
The innovative use of the Choquet Integral has enabled the researchers to identify key health factors critical for weight loss, improving the precision and efficiency of their study methods over traditional approaches. This method not only minimises errors but also enhances the extraction of useful information by concentrating on the most pertinent data. The Choquet Integral utilises a fuzzy measure, which unlike traditional metrics that provide definite values such as size or length, can manage uncertainty and overlaps. This allows for a more adaptable evaluation of complex, interrelated data sets.
Matthew Fried, the leading author of the paper and a Ph.D. student at Katz School under the mentorship of Dr. Honggang Wang, explained, “We believe this methodology could pave the way for more efficient and accurate health data analysis, ultimately contributing to better health outcomes and advancing the fight against obesity.”
The methodology was rigorously tested across four distinct datasets, including random numbers, fabricated data, standard heart data from the UC Irvine Libraries, and National Institutes of Health (NIH) health data. The technique proved effective in differentiating between actual data and noise, underscoring its suitability in modelling interactive features and measuring variables like insulin and glucose levels, LDL (bad cholesterol), HDL (good cholesterol), height, and more.
Dr. Honggang Wang, co-author and chair of the Graduate Department of Computer Science and Engineering at Katz School, stated, “We studied how different health factors affect each other, whether positively or negatively, using this special mathematical approach. This method helped us understand more clearly which biological factors are most important for weight loss.”
This cross-disciplinary application of the Choquet Integral to health data analysis not only enhances the efficiency of machine learning models by selecting reduced versions of power sets but also has the potential to transform weight loss studies and other areas in the biomedical field.
Hua Fang, another co-author and professor of computer & information science at the University of Massachusetts Dartmouth and UMass Chan Medical School, noted, “The benefits of this technique extend beyond weight loss studies. It has broad potential applications in various biomedical fields where analysing complex inter-variable relationships is crucial.” This marks a significant step forward in the use of mathematical models to enhance biomedical research and health care outcomes.
