Graph neural networks (GNNs) have recently emerged as state-of-the-art collaborative filtering (CF) solution. A fundamental challenge of CF is to distill negative signals from the implicit feedback, but sampling in GNN-based has been largely unexplored. In this work, we propose study by leveraging both user-item graph structure and GNNs' aggregation process. We present MixGCF method---a general plugin that can be directly used train recommender systems. MixGCF, rather than raw negatives...

Graph representation learning has been extensively studied in recent years, which sampling is a critical point. Prior arts usually focus on positive node pairs, while the strategy for negative left insufficiently explored. To bridge gap, we systematically analyze role of from perspectives both objective and risk, theoretically demonstrating that as important determining optimization resulted variance. best our knowledge, are first to derive theory quantify nice distribution pn(u|v) ∝...

The intermittency and uncertainty of the renewable energy deteriorate stability microgrids. In this article, we focus on a grid-connected microgrid with wind power battery storage system (BESS). electricity load is satisfied by from turbine, BESS, grid, together. purpose to reduce fluctuation grid charging or discharging BESS dynamically. A Markov chain used depict randomness measured variance power. Since measurement quadratic nonadditive, dynamic optimization problem does not fit standard...

Graph neural network-based recommendation systems are blossoming recently, and its core component is aggregation methods that determine neighbor embedding learning. Prior arts usually focus on how to aggregate information from the perspective of spatial structure information, but temporal about neighbors left insufficiently explored.

Abstract Energy efficiency of data centers (DCs) is great concern due to their large amount energy consumption and the foreseeable growth in demand digital services future. The past decade witnessed improvements DCs from an extensive margin—a shift small large, more efficient DCs. Improvements intensive margin, that is, operation, would be critical limiting environmental impact upcoming period. Machine learning algorithms have advantages optimizing operation improve as they shown potential...

Graph representation learning has been extensively studied in recent years. Despite its potential generating continuous embeddings for various networks, both the effectiveness and efficiency to infer high-quality representations toward large corpus of nodes are still challenging. Sampling is a critical point achieve performance goals. Prior arts usually focus on sampling positive node pairs, while strategy negative left insufficiently explored. To bridge gap, we systematically analyze role...

Vision-language pre-training (VLP) methods are blossoming recently, and its crucial goal is to jointly learn visual textual features via a transformer-based architecture, demonstrating promising improvements on variety of vision-language tasks. Prior arts usually focus how align features, but strategies for improving the robustness model speeding up convergence left insufficiently explored.In this paper, we propose novel method ViLTA, comprising two components further facilitate fine-grained...

Abstract In the past ten years, deep learning has achieved remarkable results in area of natural image segmentation, and gradually turned to field medical segmentation. The precise segmentation spine images can be used for early screening spondylopathy, which is convenient detection treatment patients. Aiming at by U-Net network, structure will lead large model calculation, network overfitting, size, noise information other issues. This paper introduces a new method based on spatial pyramid...

The intrinsic randomness of renewable energy has a negative impact on the safety power grid. In this paper, we aim at decreasing large fluctuations output from wind farm integrated with battery storage system (BESS), so as to improve stability and quality system. control method is dynamically charge or discharge BESS, coordinated limited curtailment. fluctuation total measured by variance, which reflects risk difficulty that dynamic optimization problem does not meet requirement standard...

With high penetration of renewable generation, power system operations need additional ramping capability to follow the change generation. However, without proper unit commitment in forward energy scheduling process, may not have sufficient real time operation when generation varies while load is sharply pulling up or down. This paper analyzes needs incorporate uncertainty and proposes a ramp enhanced model for day-ahead scheduling. Simulations are performed on WSCC179 with manufactured...

Due to the characteristics of intermittency and uncertainty wind power, drastic power fluctuations have negative impacts on safety stability systems. In this paper, we focus a farm with battery energy storage system (BESS). The objective is reduce fluctuation total output by dynamically scheduling charging discharging BESS. randomness characterized periodic Markov chain from long-term viewpoint. variance chosen as indicator fluctuations. However, since function quadratic non-additive,...

The multi researches and experiments show that the future highway traffic accident situation is shown by prediction. In paper, support vector regression trained genetic algorithm presented in method, used to train parameters of regression. Firstly, function introduced, are optimized algorithm. computation results between G-SVR SVR can indicate prediction ability for accidents better than absolutely.

The rotation of electron vortex beams (EVBs) presents a complex interplay the Gouy phase characterizing free-space behavior and Landau states or Larmor observed in magnetic fields. Despite being studied separately, these phenomena manifest within single beam during its propagation fields, lacking comprehensive description. We address this by utilizing exact solutions relativistic paraxial equation termed "paraxial modes". modes describe quantum EVBs Our study angles demonstrates consistency...

Container terminals play a crucial role in global logistics and trade, with gate-in operations significantly impacting overall terminal efficiency cargo turnover speed. This paper analyzes series of problems caused by the randomness arrival export containers at container yard, including wastage yard space, excessive waiting time for external trucks, conflicts other production operations. To address these issues, method based on decomposed ensemble framework is proposed to predict short-term...

In this paper, we propose a new optimization method to simultaneously maximize the average return and minimize reward variance of stochastic dynamic system. This problem cannot be formulated as standard Markov decision process (MDP) since criterion is combined metric with mean variance. Traditional methods, such programming, are not valid. We resort sensitivity-based theory solve problem. derive performance difference formula which quantifies mean-variance metrics under any two different...

Predicting accurate location of Protein Subcellular is conductive to acknowledging the function protein and finding cancer biomarkers. Unfortunately, many experimental approaches for classifying subcellular are still high-cost time-consuming. However, deep convolutional neural network has achieved significant advances in fields, such as image classification, object detection segmentation, it's driving us use classify images. Because unavoidable differences between bioimages natural images,...

The simulation of high accuracy three-dimension (3D) scene optical field radiation distribution can provide the input for camera design, optimization key parameters and testing various imaging models. It benefit reducing strong coupling between models simulation. However, computation is extremely large non-optimization computing method can't performed efficiently. Therefore, a study was carried out from algorithm using high-performance platform to accelerate operation speed. On one hand,...