Extracting invariant representations in unlabeled electrocardiogram (ECG) signals is a challenge for deep neural networks (DNNs). Contrastive learning is a promising method for unsupervised learning. However, it should improve its robustness to noise and learn the spatiotemporal and semantic representations of categories, just like cardiologists. This article proposes a patient-level adversarial spatiotemporal contrastive learning (ASTCL) framework, which includes ECG augmentations, an adversarial module, and a spatiotemporal contrastive module. Based on the ECG noise attributes, two distinct but effective ECG augmentations, ECG noise enhancement, and ECG noise denoising, are introduced. These methods are beneficial for ASTCL to enhance the robustness of the DNN to noise. This article proposes a self-supervised task to increase the antiperturbation ability. This task is represented as a game between the discriminator and encoder in the adversarial module, which pulls the extracted representations into the shared distribution between the positive pairs to discard the perturbation representations and learn the invariant representations. The spatiotemporal contrastive module combines spatiotemporal prediction and patient discrimination to learn the spatiotemporal and semantic representations of categories. To learn category representations effectively, this article only uses patient-level positive pairs and alternately uses the predictor and the stop-gradient to avoid model collapse. To verify the effectiveness of the proposed method, various groups of experiments are conducted on four ECG benchmark datasets and one clinical dataset compared with the state-of-the-art methods. Experimental results showed that the proposed method outperforms the state-of-the-art methods.

Purpose: This paper aims to find optimal emission reduction investment strategies for the manufacturer and examine the effects of carbon cap-and-trade policy and uncertain low-carbon preferences on emission reduction investment strategies.Design/methodology/approach: This paper studied a supply chain consisting of one manufacturer and one retailer, in which the manufacturer is responsible for emission reduction investment. The manufacturer has two emission reduction investment strategies: (1) invest in traditional emission reduction technologies only in the production process and (2) increase investment in smart supply chain technologies in the use process. Then, three different Stackelberg game models are developed to explore the benefits of the manufacturer in different cases. Finally, this paper coordinates between the manufacturer and the retailer by developing a revenue-sharing contract.Findings: The manufacturer's optimal emission reduction strategy is dynamic. When consumers' low-carbon preferences are low and the government implements a carbon cap-and-trade policy, the manufacturer can obtain the highest profit by increasing the emission reduction investment in the use process. The carbon cap-and-trade policy can encourage the manufacturer to reduce emissions only when the initial carbon emission is low. The emission reduction, order quantity and the manufacturer's profit increase with the consumers' low-carbon preferences. And the manufacturer can adjust the emission reduction investment according to the emission reduction cost coefficient in two processes.Originality/value: This paper considers the investment of emission reduction technologies in different processes and provides theoretical guidance for manufacturers to make a low-carbon transformation. Furthermore, the paper provides suggestions for governments to effectively implement carbon cap-and-trade policy.

The effects of greenhouse gas (GHG) on the weather has caused ever-increasing disasters around the world. Many countries are making significant efforts to reduce GHG emissions in all industrial sectors. In this study, a green vehicle routing problem (GVRP) has been formulated as a nonlinear integer programming problem to minimize GHG emissions, considering various realistic factors that include three-dimensional customer locations, gravity, vehicle speed, vehicle operating time, vehicle capacity, rolling resistance, air density, road grade and inertia. Lagrangian relaxation has been introduced to propose a simple solution method. In contrast to traditional vehicle routing problems, the vehicle speed, vehicle weight, and road grade between two customer locations are also determined along with vehicle routes. The computational results demonstrate the effectiveness and efficiency of the proposed solution method.