Feature Importance of Stabilised Rammed Earth Components Affecting the Compressive Strength Calculated with Explainable Artificial Intelligence Tools

Abstract

Cement-stabilized rammed earth (CSRE) is a sustainable construction material. The use of it allows for economizing on the cost of a structure. These two properties of CSRE are based on the fact that the soil used for the rammed mixture is usually dug close to the construction site, so it has random characteristics. That is the reason for the lack of widely accepted prescriptions for CSRE mixture, which could ascertain high enough compressive strength. Therefore, assessing which components of CSRE have the highest impact on its compressive strength becomes an important issue. There are three machine learning regression tools, i.e., artificial neural networks, decision tree, and random forest, used for predicting the compressive strength based on the relative content of CSRE composites (clay, silt, sand, gravel, cement, and water content). The database consisted of 434 samples of CSRE, which were prepared and crushed for testing purposes. Relatively low prediction errors of aforementioned models allowed for the use of explainable artificial intelligence tools (drop-out loss, mean squared error reduction, accumulated local effect) to rank the influence of the ingredients on the dependent variable—the compressive strength. Consistent results from all above-mentioned methods are discussed and compared to some statistical analysis of selected features. This innovative approach, helpful in designing the construction material is a solid base for reliable conclusions.

Designing the Composition of Cement-Stabilized Rammed Earth with the Association Analysis Application

“…The created database is large enough to apply machine learning tools. Some of them have already been applied by the authors in [ 3 , 6 , 63 ]. The compressive strength is predicted, and an impact of CSRE ingredients on the compressive strength is searched there.…”

“…A significant problem of using CSRE for construction is the uncertainty of the obtained mechanical properties of the material, in particular, the compressive strength [ 3 ]. It is influenced by several properties of the soil–cement mixture, such as soil grain size [ 3 , 4 , 5 , 6 , 7 , 8 , 9 ], soil mineral composition [ 10 , 11 , 12 ], amount and type of the cement used [ 3 , 5 , 9 , 13 , 14 ], dry density [ 3 , 4 , 6 , 9 ] and the mixture’s moisture [ 3 , 9 , 15 , 16 ], curing time [ 17 ]. Moreover, the method and energy of compacting of soil layers, their height, and homogeneity also have an impact on the mechanical strength [ 18 , 19 ].…”

Analysis of Environmental Factors Affecting the Atmospheric Corrosion Rate of Low-Alloy Steel Using Random Forest-Based Models

“…The importance indicates how much the feature affects the prediction results. It is often used as an efficient tool in the feature screening and correlation analysis [ 36 ].…”

Prediction of Mechanical Properties by Artificial Neural Networks to Characterize the Plastic Behavior of Aluminum Alloys

“…One of the most important is that they usually carry out such complex processing that involves thousands of operations, so it is not possible to follow, step by step, the reasoning that led them to draw their conclusions [ 41 , 57 ]. Nevertheless, in small networks, by simulation or by studying synaptic weights, it is possible to know, at least, which input variables have been relevant in decision-making [ 58 , 59 , 60 ].…”