An Uncertainty Measure for Prediction of Non-Gaussian Process Surrogates.

Model management is an essential component in data-driven surrogate-assisted evolutionary optimization. In model management, the solutions with a large degree of uncertainty in approximation play an important role. They can strengthen the exploration ability of algorithms and improve the accuracy of surrogates. However, there is no theoretical method to measure the uncertainty of prediction of Non-Gaussian process surrogates. To address this issue, this article proposes a method to measure the uncertainty. In this method, a stationary random field with a known zero mean is used to measure the uncertainty of prediction of Non-Gaussian process surrogates. Based on experimental analyses, this method is able to measure the uncertainty of prediction of Non-Gaussian process surrogates. The method's effectiveness is demonstrated on a set of benchmark problems in single surrogate and ensemble surrogates cases.

On Nonstationary Gaussian Process Model for Solving Data-Driven Optimization Problems.

In data-driven evolutionary optimization, most existing Gaussian processes (GPs)-assisted evolutionary algorithms (EAs) adopt stationary GPs (SGPs) as surrogate models, which might be insufficient for solving most optimization problems. This article finds that GPs in the optimization problems are nonstationary with great probability. We propose to employ a nonstationary GP (NSGP) surrogate model for data-driven evolutionary optimization, where the mean of the NSGP is allowed to vary with the decision variables, while its residue variance follows an SGP. In this article, the nonstationarity of GPs in the tested functions is theoretically analyzed. In addition, this article constructs an NSGP where the SGP is a degenerate case. Performance comparisons of the NSGP with the SGP and the NSGP-assisted EA (NSGP-MAEA) with the SGP-assisted EA (SGP-MAEA) are carried out on a set of benchmark problems and an antenna design problem. These comparison results demonstrate the competitiveness of the NSGP model.

Association of systemic inflammation and body mass index with survival in patients with resectable gastric or gastroesophageal junction adenocarcinomas.

Objective: The systemic inflammation index and body mass index (BMI) are easily accessible markers that can predict mortality. However, the prognostic value of the combined use of these two markers remains unclear. The goal of this study was therefore to evaluate the association of these markers with outcomes based on a large cohort of patients with gastric cancer. Methods: A total of 2,542 consecutive patients undergoing radical surgery for gastric or gastroesophageal junction adenocarcinoma between 2009 and 2014 were included. Systemic inflammation was quantified by the preoperative neutrophil-to-lymphocyte ratio (NLR). High systemic inflammation was defined as NLR ≥ 3, and underweight was defined as BMI < 18.5 kg/m2. Results: Among 2,542 patients, NLR ≥ 3 and underweight were common [627 (25%) and 349 (14%), respectively]. In the entire cohort, NLR ≥ 3 or underweight independently predicted overall survival (OS) [hazard ratio (HR): 1.236, 95% confidence interval (95% CI): 1.069-1.430; and HR: 1.600, 95% CI: 1.350-1.897, respectively] and recurrence-free survival (RFS) (HR: 1.230, 95% CI: 1.054-1.434; and HR: 1.658, 95% CI: 1.389-1.979, respectively). Patients with both NLR ≥ 3 and underweight (vs. neither) had much worse OS (HR: 2.445, 95% CI: 1.853-3.225) and RFS (HR: 2.405, 95% CI: 1.802-3.209). Furthermore, we observed similar results in subgroup analyses according to pathological stage, age, and postoperative chemotherapy. Conclusions: Our results showed that preoperative elevated NLR and decreased BMI had a significant negative effect on survival. Underweight combined with severe inflammation could enhance prognostication. Taking active therapeutic measures to reduce inflammation and increase nutrition may help improve outcomes.