Response-balanced sampling shifts but does not improve slope stability prediction accuracy
Response-balanced sampling relocates rather than improves accuracy in slope-stability surrogates under a fixed simulation budget
Computational Engineering, Finance, and Science
Summary
When predicting how likely slopes are to collapse, machine learning models need careful training with data from simulations. The authors tested a method that tries to balance the training data according to the predicted risk levels instead of sampling randomly. They found this balancing approach does not make the predictions more accurate overall; instead, it just moves improved accuracy from one risk level range to another. Even when the ideal balanced data is provided, the usual random sampling method performs just as well for decision-making purposes.
What this means in practice
- •For geotechnical engineers: Choose sampling methods for building machine-learning models to predict slope failure risks without expecting automatic accuracy improvements from response-balancing strategies.
- •For machine learning practitioners in geosciences: Allocate simulation budgets with Latin hypercube sampling confidently when building surrogate models for physical analysis, knowing balancing may not improve key prediction accuracy.
Authors
Iancu-Bogdan Teodoru, Alessia Amelio, Mircea Aniculăesi, Irina Lungu
Abstract
Machine-learning surrogates of slope stability learn factors of safety (FOS) that only physics-based analyses produce, so analyses are the budget to allocate. Because Latin hypercube sampling (LHS) of the parameters covers the FOS unevenly, a natural remedy is to choose part of the training points to even out the FOS distribution. We test this response-balanced sampling at a fixed budget. A random-forest screener trained on a 300-point LHS seed predicts the FOS of 200,000 candidates, and 300 further limit-equilibrium analyses are allocated to even out the FOS over the selection range, $[0.5, 6]$. The design is compared with 300 further LHS analyses over five realizations of the added points, on two test sets, under two independently tuned tree-ensemble families, with ablations throughout; an oracle drawing from 22,000 solved analyses meets the intended density exactly. The primary criterion is the log-scale root-mean-square error (log-RMSE) on the decision range, FOS $\in [0.5, 2)$, against continued LHS. Response-balanced sampling does not systematically improve accuracy; it relocates accuracy along the FOS axis, consistently across realizations, from $[1, 2)$ to $[0.5, 1)$, following the training density actually achieved, even when the intended density is met exactly. On the decision range no screener-selected design shows a demonstrated gain, and with known bin membership that verdict stands, in this design. Within the selection range, LHS gives $[0.5, 1)$ 11-12% of the points against a log-uniform target's 27.9%, but $[1, 2)$ already 35-37%: the region that is rare need not be where decisions are made.