Designing Social Care Policies in the Long Term: A Study Using Regression, Clustering and Backpropagation Neural Nets

Authors: Sotirios Raptis

Abstract:

Linking social needs to social classes using different criteria may lead to social services misuse. The paper discusses using ML and Neural Networks (NNs) in linking public services in Scotland in the long term and advocates, this can result in a reduction of the services cost connecting resources needed in groups for similar services. The paper combines typical regression models with clustering and cross-correlation as complementary constituents to predict the demand. Insurance companies and public policymakers can pack linked services such as those offered to the elderly or to low-income people in the longer term. The work is based on public data from 22 services offered by Public Health Services (PHS) Scotland and from the Scottish Government (SG) from 1981 to 2019 that are broken into 110 years series called factors and uses Linear Regression (LR), Autoregression (ARMA) and 3 types of back-propagation (BP) Neural Networks (BPNN) to link them under specific conditions. Relationships found were between smoking related healthcare provision, mental health-related health services, and epidemiological weight in Primary 1(Education) Body Mass Index (BMI) in children. Primary component analysis (PCA) found 11 significant factors while C-Means (CM) clustering gave 5 major factors clusters.

Keywords: Probability, cohorts, data frames, services, prediction.

References:

[1] Simon Bottery . https : //www.kingsfund.org.uk/about − us/whos − who/simon − bottery?page = 2 . The King’s Fund
[2] Public Health Scotland (2020). Data and intelligence. A – Z Subject Index. https : //www.isdscotland.org/A−to−Z−index/index.asp
[3] Scottish Government (2019). Statistics Service Health and Social Care Data. https : //statistics.gov.scot/datahome
[4] Vittorio Lippi (2019). Incremental Principal Component Analysis: Exact implementation and continuity corrections. arXiv: 1901:07922v2; stat:ML; 13May2019. https : //arxiv.org/pdf/1901.07922.pdf
[5] Ian Litchfield (2019). Can pathways of patients with long-term conditions in UK primary care? A study protocol. BMJ Open, 2018. https : //bmjopen.bmj.com/content/8/12/e019947
[6] Dimitris Bertsimas (2018), Colin Pawlowski, Ying Daisy Zhuo (2018). From Predictive Methods to Missing Data Imputation: An Optimisation Approach. Journal of Machine
[7] E.M. Mirkes (2018), T.J. Coats, J. Levesley, A.N. Gorban (2018). From Predictive Methods to Missing Data Imputation: An Optimisation Approach. Journal of Machine Learning Research, 18 (2018),1-39. http : //dx.doi.org/10.1016/j.compbiomed.2016.06.004
[8] deRooij M. (2018). Transitional modeling of experimental longitudinal data with missing values. Adv Data AnalClassif, 12,107–130. https : //link.springer.com/article/10.1007/s11634 − 015 − 0226 − 6
[9] Burd M. (2015) . The Health System Matrix 6.1: Understanding the Health Care Needs of the British Columbia Population through Population Segmentation. https : //doi.org/10.1186/s13643 − 019 − 1105 − 6
[10] Gredell Devin (2019). Comparison of Machine Learning Algorithms for Predictive Modeling of Beef Attributes Using Rapid Evaporative Ionization Mass Spectrometry (REIMS). Data. Sci Rep.,9 5721 (2019). https : //pubmed.ncbi.nlm.nih.gov/30952873/
[11] Md Saiful Islam, Md Mahmudul Hasan (2018). A Systematic Review on Healthcare Analytics: Application and Theoretical Perspective of Data Mining, Healthcare, 2018,6-54. doi : 10.3390/healthcare6020054
[12] Muge Capan (2020), Stephen Hoover, et al. (2019). Time Series Analysis for Forecasting Hospital Census: Application to the Neonatal Intensive Care Unit Multitask learning and benchmarking with clinical time series data. Appl. Clin. Inform.. 2019,7(2):275–289. https : //dx.doi.org/10.4338%2FACI − 2015 − 09 − RA − 0127
[13] Langton, J.M. (2018), Wong, S.T., Burge, F. et al. (2015). Population segments as a tool for health care performance reporting: an exploratory study in the Canadian province of British Columbia. BMC Fam Pract,21- 98(2020). https : //doi.org/10.1186/s12875 − 020 − 01141 − w
[14] Vimal Mishra (2019), MD, MMCI, Shin-Ping Tu, MD, MPH, Joseph Heim, PhD, Heather Masters, MD, Lindsey Hall, MPH, Ralph R. Clark, MD, Alan W. Dow, MD (2019). Predicting the Future: Using Simulation Modeling to Forecast Patient Flow on General Medicine Units. J. Hosp. Med.,2019,1,9-15. doi:10.12788/jhm.3081
[10]
[15] Guersel, Gueney (2019). Healthcare, uncertainty, and fuzzy logic. Digital Medicine ,2016,2,101-12. https : //www : researchgate : net = publication = 310817255Healthcareuncertaintyandfuzzylogic
[16] Deborah A.Marshall, LinaBurgos-Liz et al. (2015). Applying Dynamic Simulation Modeling Methods in Health Care Delivery Research—The SIMULATE Checklist:Report of the ISPOR Simulation Modeling Emerging Good Practices Task Force. Value in Health, volume 18,Issue 2,March 2015,143-144. https : //doi.org/10.1016/j.jval.2014.12.001
[17] D. Ben-Tovim, J. Filar, et al. (2019). Hospital Event Simulation Model: Arrivals to Discharge. 21st International Congress on Modelling and Simulation. Gold Coast,Australia. https : //www.mssanz.org.au/modsim2015/H2/bentovim.pdf
[18] Bebbington E, Furniss, D . (2015). Linear regression analysis of Hospital Episode Statistics predicts a large increase in demand for elective hand surgery in England. J. Plast. Reconstr. Aesthet. Surg, 2015, Feb,68(2),243- 51. doi:10.1016/j.bjps.2014.10.011
[19] Uematsu, H., Yamashita, K., Kunisawa, S., Otsubo, T., & Imanaka, Y. (2017). Prediction of pneumonia hospitalization in adults using health checkup data. PloS one,12(6),e0180159. https : //doi.org/10.1371/journal.pone.0180159
[20] Juang WC, Huang SJ, Huang FD, Cheng PW, Wann SR. (2017). Application of time series analysis in modelling and forecasting emergency department visits in a medical centre in Southern Taiwan. BMJ Open, 2017, Dec 1,7(11),e018628. DOI: 10.1136/bmjopen−2017−018628
[21] Harutyunyan, H., Khachatrian, H., Kale, D.C. et al. (2019). Multitask learning and benchmarking with clinical time series data. Sci .Data,6,96(2019).https : //doi.org/10.1038/s41597−019−0103−9
[22] Bui C., Pham N., Vo A., Tran A., Nguyen A., Le T. (2017). Time Series Forecasting for Healthcare Diagnosis and Prognostics with the Focus on Cardiovascular Diseases. Vo Van T.; Nguyen Le T.; https : //link.springer.com/chapter/10.1007/978 − 981 − 10 − 4361 − 1138
[23] Liew, B.X.W., Peolsson, A., Rugamer, D. et al. (2020). Clinical predictive modelling of post-surgical recovery in individuals with cervical radiculopathy: a machine learning approach. Sci.Rep,10,16782(2020). https : //doi.org/10.1038/s41598 − 020 − 73740 − 7
[24] Dunsmuir WT (2019). Dangers and uses of cross-correlation in analyzing time series in perception, performance, movement, and neuroscience: The importance of constructing transfer function autoregressive models. Behav Res Methods,2016,Jun,48(2),783-802. DOI:10.3758/s13428 − 015 − 0611 − 2
[25] Skiera, Bernd & Reiner, Jochen (2018). Regression Analysis. Homburg, Christian, Klarmann, Martin, Vomberg, Andreas (Editors). Handbook of Market Research. https : //doi : org = 10 : 1007 = 978−3−319− 05542 −8 17− 1
[26] Shivapratap Gopakumar (2016), Truyen Tran, Wei Luo, Dinh Phung, . JMIR Medical Informatics 4(3):e25 . DOI: 10.2196/medinform.5650 .
[27] Ciprian Florescu & Christian Igel (2018), RESILIENT BACKPROPAGATION (RPROP) FOR BATCHLEARNING IN TENSORFLOW. Work-shop track - ICLR 2018 . https : //openreview.net/pdf?id = r1R0o7yDz
[28] Ippoliti, R., Falavigna, G., Zanelli, C. et al. . Neural networks and hospital length of stay: an application to support healthcare management with national benchmarks and thresholds. . Cost Eff Resour Alloc 19, 67 (2021). https : //doi.org/10.1186/s12962 − 021 − 00322 − 3
[29] Yang, C., Delcher, C., Shenkman, E. et al. (2019). Expenditure variations analysis using residuals for identifying high health care utilizers in a state Medicaid program. BMC Med Inform Decis Mak, 19,131(2019). https : //doi.org/10.1186/s12911/019/0870/4
[30] Daniel J. Morgan, Bill Bame, Paul Zimand, et al. (2019). Assessment of Machine Learning vs Standard Prediction Rules for Predicting Hospital Readmissions. JAMA Netw Open,2019,Mar,2
[31] Boelaert, Julien & Ollion, Etienne. (2018). The Great Regression. Machine Learning, Econometrics, and the Future of Quantitative Social Sciences. Revue franc¸aise de sociologie. 59.
[32] Marno Verbeek . A Guide to Modern Econometrics . John Wiley & Sons. DOI10.3917/rfs.593.0475. https : //www.researchgate.net/publication/227488993 A Guide to Modern Econometrics
[33] Aitor Lewkowycz and Ethan S Dyer and Guy Gur-Ari and Jascha Sohl-dickstein and Yasaman Bahri (2020) . ICLR 2021 Conference . The large learning rate phase of deep learning . https : //arxiv.org/abs/2003.02218
[34] Liu C, Zhang X, Nguyen TT, et al. (2021). Partial least squares regression and principal component analysis: similarity and differences between two popular variable reduction approaches. General Psychiatry,2022; 35 : e100662(2021). doi: 10.1136/gpsych − 2021 − 100662
[35] Torti, F., Perrotta, D., Riani, M. et al. Assessing trimming methodologies for clustering linear regression data. Adv Data Anal Classif 13, 227–257 (2019). https://doi.org/10.1007/s11634-018-0331-4
[36] Lyall DM, Inskip HM, Mackay D, Deary IJ, McIntosh AM, Hotopf M, Kendrick T, Pell JP, Smith DJ. Low birth weight and features of neuroticism and mood disorder in 83545 participants of the UK Biobank cohort . BJPsych Open. 2016 Jan 28,2(1):38-44. DOI:10.1192/ bjpo.bp.115.002154. PMID : 27703752, PMCID : PMC4995581