Freezing temperature changes through the remediation of pb-contaminated soils using biochar: Capability analysis of tree algorithm

Biochar alters soil structure, removes lead contamination, and significantly influences the soil freezing point. The work used machine learning approaches, including Decision Tree Regression (DTR), to examine the alteration in freezing temperature (ToF) that transpired throughout the lead removal process from biochar-polluted soils. The samples utilized in this study are taken from previously carried out experiments. Numerous techniques, such as the Red-Tailed Hawk Optimization (RTHO) and the Hippopotamus Optimization (HO), are used to achieve optimization goal. It was determined by the study team that 288 samples of the dataset would be used for learning and model creation, and 72 samples would be used for model evaluation. The uncertainty analysis (U95) values for the learning and testing phases were 0.4293 and 0.4423, respectively, according to the Decision Tree Regression integrated with Hippopotamus Optimization (DTRH). The equivalent U95 values for the Decision Tree Regression integrated with Red-Tailed Hawk Optimization (DTRRTH) were 0.4858 (variance of -8.96844%) and 0.4485 (variance of -4.27751%). The evaluation of each model in terms of logical inference, rating score, and variance percentage has shown that it is reliable and trustworthy. In terms of fulfilling its stated purpose, the DTRH model shows a slight advantage over the alternative model.