TY - JOUR ID - 59612 TI - An integrated heuristic method based on piecewise regression and cluster analysis for fluctuation data (A case study on health-care: Psoriasis patients) JO - Journal of Industrial and Systems Engineering JA - JISE LA - en SN - 1735-8272 AU - Bagheri, Farnoosh AU - laari, mahsa AU - Kamranrad, Reza AU - Jalili, Majid AD - Department of Industrial Engineering, University of Science and Culture, Tehran, Iran AD - Department of Industrial Engineering, Material and Energy Research Center, Karaj, Iran Y1 - 2018 PY - 2018 VL - 11 IS - 2 SP - 175 EP - 189 KW - Piecewise regression KW - node KW - Clustering KW - Mars algorithm KW - health-care systems DO - N2 - Trend forecasting and proper understanding of the future changes is necessary for planning in health-care area.One of the problems of analytic methods is determination of the number and location of the breakpoints, especially for fluctuation data. In this area, few researches are published when number and location of the nodes are not specified.In this paper, a clustering-based method is developed to obtain the number and the location of breakpoints. We propose an appropriate piecewise regressionmodel to analyze the fluctuation data and predict trends of them.Theefficiency of proposed integrated approach is evaluated by using a simulated and real example, and results are compared with results of Mars algorithm. Comparison shows that proposed approach has less sum of square error (SSE) criterion than Mars algorithm with equall number of nods. UR - https://www.jise.ir/article_59612.html L1 - https://www.jise.ir/article_59612_4b192da3e9a86efbd7fa37fce6b4d284.pdf ER -