Herlina Jusuf
Abstract
MARS (Multivariate Adaptive Regression Spline) is one of the non-parametric regression models that employs modified recursive partitioning algorithm. This study applied MARS model in the data of Lead (Pb) exposure in the workersÂ’ hair at the petrol station in Gorontalo City, Indonesia. The study is intended to investigate the content of Lead in oneÂ’s body. The results of the study reveal that the MARS model in the Lead exposure in the workersÂ’ hair at the petrol station in Gorontalo City consists of Y = 1.93593 - 0.219669 * BF8 - 0.0220152 * BF13. Based on that MARS model, there are only three variables out of 10 assumed to affect the Lead exposure in the workersÂ’ hair at the petrol station in Gorontalo City including the working time, disease symptoms, and age.