One of the major healthcare problems that manifest significant racial disparities is obesity. In the US, research suggests that African American children have increased risks of developing the issue (Rogers et al. 2015). Data mining has been employed to investigate the patterns of obesity in children and adults (Abdullah et al., 2017; Hossain, Mahmud, Hossin, Noori, & Jahan, 2018). As a result, its techniques can be used to respond to the following question: for obesity in African American children, are environmental or genetic factors more frequently encountered? Individual data mining techniques depend on the specifics of the data analyzed (McBride, Powers, Gilder, & Philips, 2016). However, determining the usefulness of their types for the stated question is possible.
Classification is a very important technique for the question since it assists in classifying the studied factors. Indeed, obesity is associated with multiple factors that are typically described as genetic, environmental, and lifestyle ones (Abdullah et al., 2017; Guerrero et al., 2016; Sahoo et al., 2015). A project that intends to respond to the stated question would need to ensure the appropriate classification of the gathered data to ensure that the comparisons of different groups of factors can be made. Consequently, as pointed out by Abdullah et al. (2017), classification is generally a standard data mining approach that is applied to obesity data. Summarization methods which aggregate and visualize the data would also be very effective in responding to the question since they assist in determining the frequency of different factors.
Moreover, as pointed out by McBride et al. (2016), data mining is often employed for identifying associations or relationships between particular variables. Since the question is partially concerned with determining relationships between variables, such techniques could verify the connection between the studied factors and obesity. Also, due to the existence of groups within the data, cluster analysis could theoretically be useful, and anomaly detection would help a researcher in determining if the data have significant deviations. According to McBride et al. (2016), the mentioned types of techniques cover more or less every approach used in data mining, which means that no particular method is probably uniquely unfit for responding to the stated question. However, the techniques that are aimed at classification and aggregation could be considered particularly useful.
References
Abdullah, F. S., Manan, N. S. A., Ahmad, A., Wafa, S. W., Shahril, M. R., Zulaily, N.,… Ahmed, A. (2017). Data mining techniques for classification of childhood obesity among year 6 school children. In T. Herawan, R. Ghazali, N. Nawi, & M. Deris (Eds.), Recent advances on soft computing and data mining. SCDM 2016. Advances in intelligent systems and computing. (pp. 465-474). Cham, Switzerland: Springer.
Guerrero, A., Mao, C., Fuller, B., Bridges, M., Franke, T., & Kuo, A. (2016). Racial and ethnic disparities in early childhood obesity: Growth trajectories in body mass index. Journal of Racial and Ethnic Health Disparities, 3(1), 129-137. Web.
Hossain, R., Mahmud, S., Hossin, M., Noori, S., & Jahan, H. (2018). PRMT: Predicting risk factor of obesity among middle-aged people using data mining techniques. Procedia Computer Science, 132, 1068-1076. Web.
McBride, S., Powers, C., Gilder, R., & Philips, B. (2016). “Big data” and advanced analytics. In S. McBride & M. Tietze (Eds.), Nursing informatics for the advanced practice nurse. (pp. 613-642). New York, NY: Springer Publishing Company.
Rogers, R., Eagle, T. F., Sheetz, A., Woodward, A., Leibowitz, R., Song, M.,… Jackson, E. A. (2015). The relationship between childhood obesity, low socioeconomic status, and race/ethnicity: Lessons from Massachusetts. Childhood Obesity, 11(6), 691-695. Web.
Sahoo, K., Sahoo, B., Choudhury, A. K., Sofi, N. Y., Kumar, R., & Bhadoria, A. S. (2015). Childhood obesity: Causes and consequences. Journal of Family Medicine and Primary Care, 4(2), 187-192. Web.