Prediction of HPC compressive strength based on machine learning.

The performance of the six models constructed by the combination of feature importance ranking shows that the performance of the model shows a decreasing trend with the decrease of the number of input features.

also in 24,192 other papers

Secondly, although SF, W, PC and AE have less influence on compressive strength than W/B, the importance value of the features still belongs to a fairly significant range.

also in 169 other papers