Estimating Medical Waste Generation Utilizing Penalized Regression Models

Öz

Medical Waste (MW) amount that has a significant impact on health and environment is increasing as a result of industrialization as well as population density. There is a need an accurate estimation waste generation amount that will be useful information to select the appropriate disposal methods and to organize the recycling and storage. Some researchers have applied conventional statistical algorithms and many kinds of Machine Learning (ML) algorithms to predict MW amount. However, to the best of our knowledge, penalized regression methods such as Ridge, Lasso, and Elastic Net regressions have not been used to predict the MW amount. 18-years real data were obtained from İstanbul Metropolitan Municipality Department Open Data Portal with the input variables namely number of hospitals, number of health personal, number of bed available at the hospital, crude birth rate and gross domestic product per capita. 80% of the total database being used for developing the models, whereas the rest 20% were used to validate the models. In order to compare their performances, 5-fold cross-validation was applied and performance measures (MAE, RMSE and R-squared) were calculated in this study. Of the penalized regression methods, the Lasso regression provided better performance than those of other models with RMSE, MAE, and R-squared of 349.56, 596.52, 0.96, respectively, whereas the second-best Ridge regression poorer accuracy with RMSE, MAE, and R-squared 1039.091, 878.25,0.88, respectively. Thus, in our case, Lasso regression can be considered better than the Ridge regression and Elastic Net regression due to the lowest RMSE and MAE values and highest R-squared. The results reveal that the proposed Lasso regression is better than the other penalized regression models to predict the MW amount.

Anahtar Kelimeler

• medical waste
• penalized regression models
• prediction
• sustainability

Cezalandırılmış Regresyon Modelleri Kullanılarak Tıbbi Atık Üretiminin Tahmini

Öz

Sağlık ve çevre üzerinde önemli bir etkiye sahip olan Tıbbi Atık (TAT) miktarı, nüfus yoğunluğunun yanı sıra sanayileşmenin bir sonucu olarak artmaktadır. Uygun bertaraf yöntemlerinin seçilmesi, geri dönüşüm ve depolamanın düzenlenmesi için yararlı bilgiler sağlayacak doğru bir atık üretim miktarı tahminine ihtiyaç vardır. Bazı araştırmacılar MW miktarını tahmin etmek için geleneksel istatistiksel algoritmaları ve birçok Makine Öğrenimi (ML) algoritmasını uygulamıştır. Ancak, bildiğimiz kadarıyla, Ridge, Lasso ve Elastic Net regresyonları gibi cezalandırılmış regresyon yöntemleri MW miktarını tahmin etmek için kullanılmamıştır. 18 yıllık gerçek veriler, İstanbul Büyükşehir Belediyesi Başkanlığı Açık Veri Portalı'ndan hastane sayısı, sağlık personeli sayısı, hastanedeki yatak sayısı, kaba doğum oranı ve kişi başına düşen gayri safi yurtiçi hasıla girdi değişkenleri ile elde edilmiştir. Toplam veri tabanının %80'i modellerin geliştirilmesi için kullanılırken, geri kalan %20'si modellerin doğrulanması için kullanılmıştır. Performanslarını karşılaştırmak için bu çalışmada 5 kat çapraz doğrulama uygulanmış ve performans ölçütleri (MAE, RMSE ve R-kare) hesaplanmıştır. Cezalandırılmış regresyon yöntemlerinden Lasso regresyonu sırasıyla 349.56, 596.52, 0.96 RMSE, MAE ve R-kare ile diğer modellerden daha iyi performans sağlarken, ikinci en iyi Ridge regresyonu sırasıyla 1039.091, 878.25, 0.88 RMSE, MAE ve R-kare ile daha düşük doğruluk sağlamıştır. Dolayısıyla, bizim durumumuzda, Kement regresyonu, en düşük RMSE ve MAE değerleri ve en yüksek R-kare nedeniyle Ridge regresyonu ve Elastik Ağ regresyonundan daha iyi kabul edilebilir. Sonuçlar, önerilen Lasso regresyonunun MW miktarını tahmin etmek için diğer cezalandırılmış regresyon modellerinden daha iyi olduğunu ortaya koymaktadır.

Anahtar Kelimeler

• tıbbi atık
• cezalandırılmış regresyon modelleri
• tahmin
• sürdürülebilirlik

Kaynakça

1. [1] Ceylan, Z.; Bulkan, S.; Elevli, S. Prediction of medical waste generation using SVR, GM (1,1) and ARIMA models: a case study for megacity Istanbul. J Environ Health Sci Engineer. 2020, 18:687–697. https://doi.org/10.1007/s40201-020-00495-8.
2. [2] Jahandideh, S.; Jahandideh, S.; Asadabadi, E.B.; Askarian, M.; Movahedi, M.M.; Hosseini, S.; Jahandideh, M. The use of artificial neural networks and multiple linear regression to predict rate of medical waste generation. Waste Manag. 2009, 29(11):2874-9. doi: 10.1016/j.wasman.2009.06.027.
3. [3] Golbaz, S.; Nabizadeh, R.; Sajadi, H.S. Comparative study of predicting hospital solid waste generation using multiple linear regression and artificial intelligence. J Environ Health Sci Engineer. 2019, 17:41–51. https://doi.org/10.1007/s40201-018-00324-z.
4. [4] Shinee, E.; Gombojav, E et al. Healthcare Waste Management in the Capital City of Mongolia. Waste Manag. 2008, 28: 435-444.
5. [5] Nie, L.; Qiao, Z.; Wu, H. Medical Waste Management in China: A Case Study of Xinxiang. J Environ Prot Ecol. 2014, 5: 803-810. http://dx.doi.org/10.4236/jep.2014.510082.
6. [6] Uysal, F.; Tinmaz, E. Medical waste management in Trachea region of Turkey: suggested remedial action. Waste Manag Res. 2004, Oct;22(5):403-7. doi: 10.1177/0734242X04045690.
7. [7] Birpinar, M.E.; Bilgili, M.S.; Erdoğan, T. Medical waste management in Turkey: A case study of Istanbul. Waste Manag. 2009, 29(1):445-8. doi:10.1016/j.wasman.2008.03.015.
8. [8] Nguyen, X.C.; Nguyen, T.T.H.; La, D.D. et al. Development of machine learning-based models to forecast solid waste generation in residential areas: A case study from Vietnam.Resour Conserv Recycl. 2021,167:105381. https://doi.org/10.1016/j.resconrec.2020.105381.

9. [9] Bdour, A.; Altrabsheh, B.; Hadadin, N.; Al-Shareif, M. Assessment of medical wastes management practice: a case study of the northern part of Jordan. Waste Manag. 2007, 27:746–59. https://doi.org/10.1016/J.WASMAN.2006.03.004.
10. [10] Sabour, M.R.; Mohamedifard, A.; Kamalan, H. A mathematical model to predict the composition and generation of hospital wastes in Iran. Waste Manag. 2007,27:584–7. https://doi.org/10.1016/J.
11. [11] Idowu, I.; Alo, B.; Atherton, W.; Al, K.R. Profile of medical waste management in two healthcare facilities in Lagos, Nigeria: a case study. Waste Manag Res. 2013, 31:494–501. https://doi.org/10.1177/0734242X13479429.
12. [12] Al-Khatib, I.A.; Abu, Fkhidah. I.; Khatib, J.I.; Kontogianni, S. Implementation of a multi-variable regression analysis in the assessment of the generation rate and composition of hospital solid waste for the design of a sustainable management system in developing countries. Waste Manag Res. 2016, 34:225–34. https://doi.org/10.1177/0734242X15622813.
13. [13] Çetinkaya, A.Y.; Kuzu, S.L.; Demir,A. Medical waste management in a mid-populated Turkish city and development of medical waste prediction model. Environ Dev Sustain. 2020, 22:6233–6244. https://doi.org/10.1007/s10668-019-00474-6.
14. [14] Chauhan, A.; Singh, A. An ARIMA model for the forecasting of healthcare waste generation in the Garhwal region of Uttarakhand. India Int J Serv Oper Informatics. 2017, 8:352. https://doi.org/10.1504/ijsoi.2017.086587
15. [15] Karpušenkaitė, A.; Ruzgas, T.; Denafas, G. Forecasting medical waste generation using short and extra short datasets: Case study of Lithuania. Waste Manag Res. 2016, 34(4):378-87. doi: 10.1177/0734242X16628977
16. [16] Thakur, V.; Ramesh, A. Analyzing composition and generation rates of biomedical waste in selected hospitals of Uttarakhand, India. J Mater Cycles Waste Manag. 2018,20:877–90. https://doi.org/10.1007/s10163-017-0648-7.
17. [17] Dissanayaka, D.M.S.H.; Vasanthapriyan, S. Forecast municipal solid waste generation in Sri Lanka. In 2019 International Conference on Advancements in Computing, Sri Lanka ,5-7 December 2019 ;210-215. doi: 10.1109/ICAC49085.2019.9103421.
18. [18] Meleko, A.; Adane, A. Assessment of Health Care Waste Generation Rate and Evaluation of its Management System in Mizan Tepi University Teaching Hospital (MTUTH), Bench Maji Zone, South West Ethiopia. Ann Rev Resear.2018, 1: 555566. http://dx.doi.org/10.19080/ARR.2018.01.555566.
19. [19] Karpušenkaitė, A.; Ruzgas, T.; Denafas, G. Time-series-based hybrid mathematical modelling method adapted to forecast automotive and medical waste generation: Case study of Lithuania. Waste Manag Res. 2018, 36: 454 - 462. [20] Papacharalampous, G.; Tyralis, H.; Koutsoyiannis, D. Univariate time series forecasting of temperature and precipitation with a focus on machine learning algorithms: A multiple-case study from Greece. Water Resour Manag. 2018, 32:5207–5239. http://dx.doi.org/10.1007/s11269-018-2155-6.
20. [21] Pavlyshenko, B. M.Machine-Learning Models for Sales Time Series Forecasting. Data. 2019,4(1):1–11. https://doi.org/10.3390/data4010015.
21. [22] Birpinar, M.E.; Bilgili, M.S.; Erdoğan, T. Medical waste management in Turkey: A case study of Istanbul. Waste Manag. 2009, 29(1):445-8. doi:10.1016/j.wasman.2008.03.015.
22. [23] Musarrat Ijaz, Zahid Asghar & Asma Gul (2021) Ensemble of penalized logistic models for classification of high-dimensional data, Communications in Statistics- Simulation and Computation, 50:7, 2072-2088, DOI: 10.1080/03610918.2019.1595647
23. [24] Buyrukoğlu, S. & Yılmaz, Y. (2021). An Approach for Airfare Prices Analysis with Penalized Regression Methods . Veri Bilimi , 4 (2) , 57-61 .
24. [25] Greenwood CJ, Youssef GJ, Letcher P, Macdonald JA, Hagg LJ, Sanson A, et al. (2020) A comparison of penalised regression methods for informing the selection of predictive markers. PLoS ONE 15(11): e0242730. https://doi.org/10.1371/ journal.pone.024273
25. [26] Rendall R, Pereira AC, Reis MS. Advanced predictive methods for wine age prediction: Part I - A comparison study of single-block regression approaches based on variable selection, penalized regression, latent variables and tree-based ensemble methods. Talanta. 2017 Aug 15;171:341-350. doi: 10.1016/j.talanta.2016.10.062. Epub 2016 Nov 9. PMID: 28551149.
26. [27] Tütmez, B. (2020). Air Quality Assessment by Statistical Learning-Based Regularization . Çukurova Üniversitesi Mühendislik-Mimarlık Fakültesi Dergisi , 35 (2) , 271-278 . DOI: 10.21605/cukurovaummfd.792412
27. [28] İstanbul Metropolitan Municipality Department Open Data Portal. Available online:https://data.ibb.gov.tr/en/dataset/31d85b21-32a9-4270-95d9-1712a6567ea/resource/50036dfd-aea5-4f06-832f-f7020fdaaa5a/download/ilce-yl-ve-atk-turu-baznda-atk-miktar-2021.xlsx (Accessed 10 July 2022)
28. [29] The official website of Turkish Statistics Institute. Available online: https://biruni.tuik.gov.tr/medas/ (accessed 10 July 2022)
29. [30] Ramírez-Gallego, S.; Krawczyk, B.; García, S.; Woźniak, M.; Herrera, F. A survey on data preprocessing for data stream mining: Current status and future directions. Neurocomputing. 2017,239:39-57. https://doi.org/10.1016/j.neucom.2017.01.078.
30. [31] Kwak, S.K.; Kim, J.H. Statistical data preparation: management of missing values and outliers. Korean J of anesthesiology. 2017, 70(4):407. doi: 10.4097/kjae.2017.70.4.407.
31. [32] James, G., Witten, D., Hastie, T., & Tibshirani, R. “An introduction to statistical learning”, Vol. 112, p. 18,. New York: springer, 2013.
32. [33] Zhang, Z., Lai, Z., Xu, Y., Shao, L., Wu, J., & Xie, G. S. “Discriminative elastic-net regularized linear regression”. IEEE Transactions on Image Processing, 26(3), 1466-1481, 2017
33. [34] Tesfahun, E.; Kumie, A.; Beyene, A. Developing models for the prediction of hospital healthcare waste generation rate. Waste Manag Res. 2016, Jan;34(1):75-80. doi: 10.1177/0734242X15607422.