Reliability Analysis of Pile Foundation Using Soft Computing Techniques: A Comparative Study

Uncertainty and variability are inherent to pile design consequently, there have been considerable researches in quantifying the reliability or probability of failure structures. This paper aims at examining comparing applicability adaptability Minimax Probability Machine Regression (MPMR), Emotional Neural Network (ENN), Group Method Data Handling (GMDH), Adaptive Neuro-Fuzzy Inference System (ANFIS) analysis embedded cohesionless soil proposes an AI-based prediction method for bearing capacity foundation. To ascertain homogeneity distribution datasets, Mann–Whitney U (M–W) Anderson–Darling (AD) tests carried out, respectively. The performance developed soft computing models is ascertained using various statistical parameters. A comparative study implemented among indices proposed by employing First Order Second Moment (FOSM). results FOSM showed that ANFIS approach outperformed other ENN worst performing model. value R2 all close 1. best RMSE achieved training phase model (0 2.13 testing) poorest (2.03 31.24 Based on experimental indices, found be very computed from original data.