Rezumat/Abstract. The resilient modulus (MR) of subgrade soil is crucial in pavement design, as it significantly affects its structural performance. However, Traditional methods, aimed at estimating this parameter, are characterized by inefficiency, time consumption, and high costs. This study introduces a novel alternative model using ten advanced machine-learning techniques including Deep Neural Network (DNN), Extreme Learning Machine (ELM), Support Vector Regression (SVR), LASSO regression (LASSO), Random Forest (RF), Ridge Regression (Ridge), Partial Least Square Regression (PLSR), Stepwise Regression (Stepwise), Kernel Ridge (KRidge), and Least Square Regression (LSR), to predict the resilient modulus (MR). The model is trained on a comprehensive dataset comprising 891 repeated load triaxial tests, and it considers nine pertinent factors as input parameters based on literature suggestions. Evaluating the efficacy of the machine-learning methods reveals that the Deep Neural Network (DNN) model outperforms others in accuracy. Subsequently, a user-friendly graphical interface called "ResiMod2024" based on the DNN model is developed to streamline the estimation process of resilient modulus, offering significant time and cost savings for researchers and civil engineers.

Cuvinte cheie/Key words: resilient modulus; machine learning; deep neural network; K-fold cross-validation approach; repeated load triaxial tests

Text integral/Full text