Relevance vector machine-based response surface for slope reliability analysis

Abstract

A new computing method is proposed for reliable analysis. The limit state function is implicit and nonlinear in reliability analysis of slopes and is difficult to apply by traditional reliability methods, especially in large-scale project applications. Relevance vector machines (RVMs) are capable of approximating the limit state function without the need for additional assumptions regarding the function form, as opposed to traditional polynomial response surfaces. RVMs were adapted to obtain the limit state function. We propose an RVM-based response surface method combined with the first-order reliability method for slope reliability analysis and describe its step-by-step implementation. The reliability index obtained from the proposed method shows excellent agreement with traditional response surface method results. Copyright © 2017 John Wiley & Sons, Ltd.