Scholarlink

This study presents a CatBoost regression model to predict the compressive strength of concrete doped with waste plastic. Input variables included % plastic waste (0–10%), % fly ash (0–10%), and curing days (3–90). The model achieved near-perfect training performance (R2 = 1.0, RMSE = 0.0003 MPa) but showed reduced accuracy on the test set (R2 = 0.762, RMSE = 4.03 MPa), indicating overfitting. Analysis revealed that increasing plastic waste reduces strength, especially beyond 4–6%, while adding 10% fly ash consistently improves strength, peaking at 41.72 MPa after 90 days. SHAP analysis confirmed % plastic waste and curing days as the dominant predictors. The model offers an effective and interpretable tool for designing sustainable concrete mixes that balance environmental and mechanical performance.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Development of a CatBoost-Based Predictive Model for Estimating the Compressive Strength of Waste Plastic-Doped Concrete

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Sameer Sen,
  • Anish Kumar,
  • Sanjeev Sinha

摘要

This study presents a CatBoost regression model to predict the compressive strength of concrete doped with waste plastic. Input variables included % plastic waste (0–10%), % fly ash (0–10%), and curing days (3–90). The model achieved near-perfect training performance (R2 = 1.0, RMSE = 0.0003 MPa) but showed reduced accuracy on the test set (R2 = 0.762, RMSE = 4.03 MPa), indicating overfitting. Analysis revealed that increasing plastic waste reduces strength, especially beyond 4–6%, while adding 10% fly ash consistently improves strength, peaking at 41.72 MPa after 90 days. SHAP analysis confirmed % plastic waste and curing days as the dominant predictors. The model offers an effective and interpretable tool for designing sustainable concrete mixes that balance environmental and mechanical performance.