Article Preview
Top2. Background
The objective of this paper is to discuss and demonstrate Machine Learning applications in public health. The article targets a universal audience because taking care of patients and producing a better outcome is a multidisciplinary initiative consisting of prognostics, diagnostics, treatment, and recovery. Practically, every aspect of today’s healthcare is subject to intelligent automation, alternatively known as Artificial Intelligence. AI-related jobs are at the highest growth rate among all professions, “With an average salary base of $146,085 and a whopping 344% growth in job postings.” (Indeed Blog, 2019).
Medical sciences are desperately looking into strengthening the after-the-fact treatment paradigm with proactive practices. A new era of health policies is asking for preventive care and active treatment plans. The Centers for Medicare and Medicaid Services (CMS) is promoting the use of AI for cost reduction, preserving patient safety, and a better outcome (Burd, Brown, Puri, & Sanghavi, 2017). The primary focus of CMS is “Better Care and a Healthier Population.” Keeping the national health policy into perspective, the focus of this research is to illustrate an application of Machine Learning into predicting the prevalence of hypertension using a large-scale health survey data from the New York City Community Health Service (CHS).
2.1 Data
The data was obtained from the New York City Community Health Survey database. “The New York City Community Health Survey (CHS) is a telephone survey conducted annually by the DOHMH, Division of Epidemiology, Bureau of Epidemiology Services. CHS provides robust data on the health of New Yorkers, including neighborhood, borough, and citywide estimates on a broad range of chronic diseases and behavioral risk factors.” (NYC, 2019). For this study, the response from the 2018 survey-year was used.
The objective was to test the application of Machine Learning with demographics, clinical, and lifestyle attributes, for a better population health prediction. In AI, data preparation and cleaning are of high importance. The initial step was to format each variable into continuous, ordinal, binary, and categorical data. In other types of Machine Learning, researchers also use text or open-ended data with numerical attributes.