Development and Comparison of Three Data Models for Predicting Diabetes Mellitus Using Risk Factors in a Nigerian Population
0
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
Objectives: This study developed and compared the performance of three widely used predictive models—logistic regression(LR), artificial neural network (ANN), and decision tree (DT)—to predict diabetes mellitus using the socio-demographic,lifestyle, and physical attributes of a population of Nigerians. Methods: We developed three predictive models using 10 inputvariables. Data preprocessing steps included the removal of missing values and outliers, min-max normalization, and featureextraction using principal component analysis. Data training and validation were accomplished using 10-fold cross-validation. Accuracy, sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and area under thereceiver operating characteristic curve (AUROC) were used as performance evaluation metrics. Analysis and model developmentwere performed in R version 3.6.1. Results: The mean age of the participants was 50.52 ± 16.14 years. The classificationaccuracy, sensitivity, specificity, PPV, and NPV for LR were, respectively, 81.31%, 84.32%, 77.24%, 72.75%, and 82.49%. Those for ANN were 98.64%, 98.37%, 99.00%, 98.61%, and 98.83%, and those for DT were 99.05%, 99.76%, 98.08%, 98.77%,and 99.82%, respectively. The best-performing and poorest-performing classifiers were DT and LR, with 99.05% and 81.31%accuracy, respectively. Similarly, the DT algorithm achieved the best AUC value (0.992) compared to ANN (0.976) and LR(0.892). Conclusions: Our study demonstrated that DT, LR, and ANN models can be used effectively for the prediction ofdiabetes mellitus in the Nigerian population based on certain risk factors. An overall comparative analysis of the modelsshowed that the DT model performed better than LR and ANN.
Abstract
Objectives: This study developed and compared the performance of three widely used predictive models—logistic regression(LR), artificial neural network (ANN), and decision tree (DT)—to predict diabetes mellitus using the socio-demographic,lifestyle, and physical attributes of a population of Nigerians. Methods: We developed three predictive models using 10 inputvariables. Data preprocessing steps included the removal of missing values and outliers, min-max normalization, and featureextraction using principal component analysis. Data training and validation were accomplished using 10-fold cross-validation. Accuracy, sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and area under thereceiver operating characteristic curve (AUROC) were used as performance evaluation metrics. Analysis and model developmentwere performed in R version 3.6.1. Results: The mean age of the participants was 50.52 ± 16.14 years. The classificationaccuracy, sensitivity, specificity, PPV, and NPV for LR were, respectively, 81.31%, 84.32%, 77.24%, 72.75%, and 82.49%. Those for ANN were 98.64%, 98.37%, 99.00%, 98.61%, and 98.83%, and those for DT were 99.05%, 99.76%, 98.08%, 98.77%,and 99.82%, respectively. The best-performing and poorest-performing classifiers were DT and LR, with 99.05% and 81.31%accuracy, respectively. Similarly, the DT algorithm achieved the best AUC value (0.992) compared to ANN (0.976) and LR(0.892). Conclusions: Our study demonstrated that DT, LR, and ANN models can be used effectively for the prediction ofdiabetes mellitus in the Nigerian population based on certain risk factors. An overall comparative analysis of the modelsshowed that the DT model performed better than LR and ANN.
참고문헌 (29)
- NCD Risk Factor Collaboration(NCD-RisC), "Worldwide trends in diabetes since 1980: a pooled analysis of 751 population-based studies with 4.4 million participants", Lancet, 387(10027), pp.1513~1530 (2016)
- International Diabetes Federation, "Diabetes atlas. 9th ed", International Diabetes Federation, ~ (2019)
- Uloko AE, "Prevalence and risk factors for diabetes mellitus in Nigeria : a systematic review and metaanalysis", Diabetes Ther, 9(3), pp.1307~1316 (2018)
- Narayan KM, "Early identification of type 2 diabetes : policy should be aligned with health systems strengthening", Diabetes Care, 34(1), pp.244~246 (2011)
- Global report on diabetes, "World Health Organization", World Health Organization, ~ (2016)
- Alebiosu OC, "Community based diabetes risk assessment in Ogun state, Nigeria(World Diabetes Foundation project 08-321)", Indian J Endocrinol Metab, 17(4), pp.653~658 (2013)
- Abbasi A, "Prediction models for risk of developing type 2 diabetes : systematic literature search and independent external validation study", BMJ, 345, pp.e5900~ (2012)
- Meng XH, "Comparison of three data mining models for predicting diabetes or prediabetes by risk factors", Kaohsiung J Med Sci, 29(2), pp.93~99 (2013)
- Habibi S, "Type 2 diabetes mellitus screening and risk factors using decision tree : results of data mining", Glob J Health Sci, 7(5), pp.304~310 (2015)
- Hajian-Tilaki K, "Sample size estimation in diagnostic test studies of biomedical informatics", J Biomed Inform, 48, pp.193~204 (2014)
- "WHO STEPS instrument question-by-question guide (core and expanded)" (2017)[URL]
- Novakovic J, "Classification performance using principal component analysis and different value of the ratio R", Int J Comput Commun Control, 6(2), pp.317~327 (2011)
- Russell S, "Artificial intelligence: a modern approach" (2010), Prentice-Hall
- Nwoye EO, "Application of artificial neural network in breast cancer classification : a comparative study", J Basic Med Sci, 2(1), pp.32~38 (2014)
- Rojas R, "Neural networks : a systematic introduction" (1996), Springer
- Sisodia D, "Prediction of diabetes using classification algorithms", Procedia Comput Sci, 132, pp.1578~1585 (2018)
- Dev VA, "Gradient boosted decision trees for lithology classification", Comput Aided Chem Eng, 47, pp.113~118 (2019)
- Lastra G, "Type 2 diabetes mellitus and hypertension : an update", Endocrinol Metab Clin North Am, 43(1), pp.103~122 (2014)
- Suastika K, "Glucose tolerance", pp.67~76 (2012), Intech Open
- Ustulin M, "Importance of family history of diabetes in computing a diabetes risk score in Korean prediabetic population", Sci Rep, 8(1), pp.15958~ (2018)
- Tillin T, "Insulin resistance and truncal obesity as important determinants of the greater incidence of diabetes in Indian Asians and African Caribbeans compared with Europeans : the Southall And Brent REvisited(SABRE)cohort", Diabetes Care, 36(2), pp.383~393 (2013)
- Leite RS, "Oral health and type 2 diabetes", Am J Med Sci, 345(4), pp.271~273 (2013)
- Lai H, "Predictive models for diabetes mellitus using machine learning techniques", BMC Endocr Disord, 19(1), pp.101~ (2019)
- El_Jerjawi NS, "Diabetes prediction using artificial neural network", Int J Adv Sci Technol, 121, pp.54~64 (2018)
- Nai-arun N, "Comparison of classifiers for the risk of diabetes prediction", Procedia Comput Sci, 69, pp.132~142 (2015)
- Wang C, "Evaluating the risk of type 2 diabetes mellitus using artificial neural network : an effective classification approach", Diabetes Res Clin Pract, 100(1), pp.111~118 (2013)
- Mohamed EI, "Predicting type 2 diabetes using an electronic nose-based artificial neural network analysis", Diabetes Nutr Metab, 15(4), pp.215~221 (2002)
- Kazemnejad A, "Comparison of artificial neural network and binary logistic regression for determination of impaired glucose tolerance", diabetes, 2010, pp.16(6), 615~620
- Li CP, "Performance comparison between logistic regression, decision trees, and multilayer perceptron in predicting peripheral neuropathy in type 2 diabetes mellitus", Chin Med J(Engl), 125(5), pp.851~857 (2012)