Bayesian optimization relies on iteratively constructing and optimizing an acquisition function. The latter turns out to be a challenging, non-convex optimization problem itself. Despite the relative importance of this step, most algorithms employ sampling- or gradient-based methods, which do not provably converge to global optima. This work investigates mixed-integer programming (MIP) as a paradigm for \emph{global} acquisition function optimization. Specifically, our Piecewise-linear Kernel Mixed Integer Quadratic Programming (PK-MIQP) formulation introduces a piecewise-linear approximation for Gaussian process kernels and admits a corresponding MIQP representation for acquisition functions. The proposed method is applicable to uncertainty-based acquisition functions for any stationary or dot-product kernel. We analyze the theoretical regret bounds of the proposed approximation, and empirically demonstrate the framework on synthetic functions, constrained benchmarks, and a hyperparameter tuning task.

Ovarian cancer is one of the most common and highly lethal gynecological malignancies, and its early diagnosis and accurate subtyping are of great clinical significance. Whole-slide images (WSI) contain rich tissue structural information required for pathological diagnosis and are often regarded as the "gold standard". However, due to their high resolution and high annotation cost, they typically only have patient-level weak labels. To address the insufficient feature representation under weak labeling and fully explore the potential histological correlations among instances, this paper proposes an ovarian cancer pathological image classification method based on Self-Supervised and multi-relational graph learning (SS-MRGL). First, self-supervised learning is employed to pre-train weakly labeled image instances, obtaining robust and discriminative compressed feature representations. Then, a multi-relational graph is constructed, and a multi-perspective joint modeling approach using Graph Convolutional Networks (GCN) and Graph Attention Networks (GAT) is introduced to capture the spatial topological and contextual semantic multi-dimensional information among instances, respectively. Finally, a gated attention mechanism is used to dynamically fuse graph-level features. Experimental results on the YN-OCPD dataset demonstrate that the proposed method can significantly improve the accuracy of ovarian cancer diagnosis under weak supervision.

Machine learning has promised to change the landscape of laboratory chemistry, with impressive results in molecular property prediction and reaction retro-synthesis. However, chemical datasets are often inaccessible to the machine learning community as they tend to require cleaning, thorough understanding of the chemistry, or are simply not available. In this paper, we introduce a novel dataset for yield prediction, providing the first-ever transient flow dataset for machine learning benchmarking, covering over 1200 process conditions. While previous datasets focus on discrete parameters, our experimental set-up allow us to sample a large number of continuous process conditions, generating new challenges for machine learning models. We focus on solvent selection, a task that is particularly difficult to model theoretically and therefore ripe for machine learning applications. We showcase benchmarking for regression algorithms, transfer-learning approaches, feature engineering, and active learning, with important applications towards solvent replacement and sustainable manufacturing.

Accurate and consistent monitoring of lake water levels is essential for understanding hydrological dynamics and climate-driven variability in remote and data-scarce regions. Satellite altimetry provides high-precision lake levels observations, but its limited spatial and temporal coverage constrain large-scale monitoring. Combining Digital Elevation Model (DEM) and remote sensing imagery offers an alternative, but the accuracy of resultant water levels is affected by the uncertainties of inherent elevation and image processing. This study proposes an XGBoost model for correcting systematic errors in inconsistent DEM-derived water levels. Twelve error-influencing parameters were incorporated, spanning lake boundary uncertainty, terrain accuracy, and area-elevation fitting errors. The results achieve a high accuracy (RMSE = 2.99 m, R2= 0.99, MAE = 0.84 m), and demonstrate robust correction performance across lakes of different sizes. We reconstructed monthly water levels time-series (2000-2021) for 965 lakes on the Qiangtang Plateau (QP) using this method. The dataset unravels divergent change trends in lake water levels on QP: a significant rise (0.12 m/a) in lakes monitored by altimetry and a slight decline (-0.028 m/a) in lakes without altimetry coverage. This highlights a potential bias when relying solely on altimetry data, as the omission of shrinking, unmonitored lakes could lead to an overestimation of regional lake expansion. We found the QP lakes are clustered into three intra-annual variation patterns, reflecting distinct hydrological and climatic influences. Uncertainties in water body extraction during frozen periods significantly reduce the accuracy of lake water levels estimations. This study provides the first plateau-scale assessment of monthly lake water levels dynamics and may offer a foundation for future research on regional hydrological changes.

Since graph neural networks (GNNs) are often vulnerable to attack, we need to know when we can trust them. We develop a computationally effective approach towards providing robust certificates for message-passing neural networks (MPNNs) using a Rectified Linear Unit (ReLU) activation function. Because our work builds on mixed-integer optimization, it encodes a wide variety of subproblems, for example it admits (i) both adding and removing edges, (ii) both global and local budgets, and (iii) both topological perturbations and feature modifications. Our key technology, topology-based bounds tightening, uses graph structure to tighten bounds. We also experiment with aggressive bounds tightening to dynamically change the optimization constraints by tightening variable bounds. To demonstrate the effectiveness of these strategies, we implement an extension to the open-source branch-and-cut solver SCIP. We test on both node and graph classification problems and consider topological attacks that both add and remove edges.

Data heterogeneity caused by Non-Independent and Identically Distributed (non-IID) data in local clients imposes limitations and challenges on the training and performance of Federated Learning. Current researches on this issue mainly focus on optimizing a single global model or developing personalized model for each client, neglecting the essential client relationships. However, despite the heterogeneity of client data, common characteristics that can be leveraged still exist. Therefore, We proposed a novel personalized federated learning approach, called pFedBEA. The method decomposes the client model into a body model (extractor) and a head model (classifier) to respectively adapt to common features and personalized attributes. Subsequently, periodically abandoning the global server aggregation and exchanging the body models among different clients, which enables local personalization while learning from multiple data sources, promoting knowledge sharing and enhancing the generalization ability of the global model. pFedBEA not only improves model performance but also reduces the number of aggregation rounds and communication time, achieving overall efficiency optimization. We conducted extensive experiments on a range of datasets, demonstrating that pFedBEA achieves higher accuracy, superior aggregation efficiency and communication efficiency.

Federated learning enables collaborative training across various clients without data exposure. However, data heterogeneity among clients may degrade system performance. The divergent training goals of servers and clients lead to performance degradation: servers aim for a global model with improved generalization across all data, whereas clients seek to develop private models tailored to their specific local data distributions. This paper introduces a novel federated learning framework named FedAMKD. FedAMKD divides federated learning into two independent entities, a local model tailored to each client's data and a global model for data aggregation and knowledge sharing. A unique aspect of FedAMKD is its adaptive mutual knowledge distillation at the local level, customized for the skewed degree of the client's data quantity. This method achieves the goal of enhancing both local and global model performance, reducing the adverse effects of data quantity-skewed heterogeneity in federated learning. Extensive experiments across diverse datasets validate FedAMKD's success in addressing challenges related to data quantity imbalances in federated learning,

Summary Collaboration between IoT end‐devices and edge servers or idle devices enables offloading of computational tasks, providing an effective solution to address inherent limitations in computational resources, storage capacity, and energy efficiency. However, IoT network openness introduces security challenges, such as privacy breaches, data security, and “free‐riding” attack. Securing computational offloading is crucial for service quality and reliability. In this article, we introduce a comprehensive approach to address these challenges. First, trust metrics are performed at the level of resource requester and resource provider respectively, and the trust level of both parties is matched to reduce the adverse effects of malicious devices or servers on computing offloading. Then, a distributed decision‐making method based on game theory is designed to improve the performance of the algorithm and avoid the single‐point‐of‐failure problem through the joint participation of multiple edge servers in decision‐making. Finally, a blockchain‐based task offloading process control strategy is used to deal with the “free‐riding” attack. The results of simulation experiments show the effectiveness of our proposed offloading decision‐making method.

Optimization over trained machine learning models has applications including: verification, minimizing neural acquisition functions, and integrating a trained surrogate into a larger decision-making problem. This paper formulates and solves optimization problems constrained by trained graph neural networks (GNNs). To circumvent the symmetry issue caused by graph isomorphism, we propose two types of symmetry-breaking constraints: one indexing a node 0 and one indexing the remaining nodes by lexicographically ordering their neighbor sets. To guarantee that adding these constraints will not remove all symmetric solutions, we construct a graph indexing algorithm and prove that the resulting graph indexing satisfies the proposed symmetry-breaking constraints. For the classical GNN architectures considered in this paper, optimizing over a GNN with a fixed graph is equivalent to optimizing over a dense neural network. Thus, we study the case where the input graph is not fixed, implying that each edge is a decision variable, and develop two mixed-integer optimization formulations. To test our symmetry-breaking strategies and optimization formulations, we consider an application in molecular design.

"Bayesian Optimization is a very effective tool for optimizing expensive black-box functions. Inspired by applications developing and characterizing reaction chemistry using droplet microfluidic reactors, we consider a novel setting where the expense of evaluating the function can increase significantly when making large input changes between iterations. We further assume we are working asynchronously, meaning we have to decide on new queries before we finish evaluating previous experiments. This paper investigates the problem and introduces 'Sequential Bayesian Optimization via Adaptive Connecting Samples' (SnAKe), which provides a solution by considering large batches of queries and preemptively building optimization paths that minimize input costs. We investigate some convergence properties and empirically show that the algorithm is able to achieve regret similar to classical Bayesian Optimization algorithms in both the synchronous and asynchronous settings, while reducing the input costs significantly. We show the method is robust to the choice of its single hyper-parameter and provide a parameter-free alternative."

Summary The combination of edge computing (EC) and the Internet of Things is a hot research topic. And security is one of the most important problems to be solved. The trust measurement of devices is an effective way to solve the security problem, and the lack of a unified trust measurement model makes the untrusted devices destroy the quality of service. Establishing a reliable trust relationship between devices can effectively improve the security of the system. A decentralized trust measurement model for devices is proposed. First, a decentralized trust measurement framework is proposed, which combines EC with blockchain technology to establish a decentralized hierarchical structure; second, the credibility of devices is measured from multilevel and multi‐attribute; finally, a feedback trust filtering mechanism is designed to filter reliable feedback information, and then the trust between devices and the comprehensive trust of devices are calculated. Experiments and analysis show that the proposed decentralized trust measurement model can effectively measure the trust degree of devices and resist various malicious feedback attacks.

River water extent is critical for understanding river discharge or its hydrological conditions. Although numerous methods have been proposed to map river water from either optical or synthetic aperture radar (SAR) remotely sensed images, uncertainties still exist broadly. In this study, we developed an image fusion method that integrates Landsat-8, Sentinel-1 and Sentinel-2 images simultaneously for river water mapping with two major steps. Firstly, a posterior probability support vector machine model was adopted to generate water probability maps from each individual image; and second, a Multi-dimensional Weighted Fusion Method (MDWFM) was developed to fuse these probability maps. Four reaches with different characteristics were selected as case study sites. High resolution aerial images were acquired and used as the reference to evaluate our results. We found the fusion process not only improves the quality of river water mapping, but also excludes the cloud interference. The fused river water maps become more reliable after the conflicts from difference images being solved by the proposed MDWFM method that contains a proportional conflict redistribution rule. The weighted root mean square difference was reduced to 0.066, and the Area Under the ROC curve reached up to 0.984. The Critical Success Index, Kappa Coefficient, and F-measure reached up to 0.810, 0.836 and 0.895, respectively. These stable and accurate river extent mapping results obtained through fusing multiple images with high spatial resolution (10 m) and short revisit interval (0.4~4.4 days) are of great significance for enriching the data and methodology of hydrological studies.

In this paper, we propose an efficient method to estimate the Weingarten map for point cloud data sampled from manifold embedded in Euclidean space. A statistical model is established to analyze the asymptotic property of the estimator. In particular, we show the convergence rate as the sample size tends to infinity. We verify the convergence rate through simulated data and apply the estimated Weingarten map to curvature estimation and point cloud simplification to multiple real data sets.

Internet of Things (IoT) equipment is usually in a harsh environment, and its security has always been a widely concerned issue. Node identity authentication is an important means to ensure its security. Traditional IoT identity authentication protocols usually rely on trusted third parties. However, many IoT environments do not allow such conditions, and are prone to single point failure. Blockchain technology with decentralization features provides a new solution for distributed IoT system. In this paper, a blockchain based multi-WSN authentication scheme for IoT is proposed. The nodes of IoT are divided into base stations, cluster head nodes and ordinary nodes according to their capability differences, which are formed to a hierarchical network. A blockchain network is constructed among different types of nodes to form a hybrid blockchain model, including local chain and public chain. In this hybrid model, nodes identity mutual authentication in various communication scenarios is realized, ordinary node identity authentication operation is accomplished by local blockchain, and cluster head node identity authentication are realized in public blockchain. The analysis of security and performance shows that the scheme has comprehensive security and better performance.

Satellite missions such as Tropical Rainfall Measurement Mission (TRMM) and Global Precipitation Mission (GPM) have been collecting a large volume of precipitation data in recent years. These data provide large scale and long-term continuous precipitation information which is useful for climate change studies. Extreme precipitation events, or so called rainstorms, are a major driven for some disasters, such as urban flooding and mountain torrents. This study proposes an efficient tool of identifying rainstorms automatically from grid based precipitation dataset. Using this tool, rainstorm events in recent 20 years in China were extracted from time series of TRMM 3B42-V7 product. The seasonal and interannual variation of rainstorms, together with their occurrences, seasonality are mapped and analyzed. Intensity of the biggest rainstorms in different regions are compared and analyzed.

Dynamic Adaptive Streaming over HTTP (DASH) has been broadly applied within most of mainstream video delivery services. At the same time, the lack of Quality of Experience (QoE) and the competition among several DASH streams have attracted significant attention. Aiming to supply better QoE for DASH streams in terms of video quality as well as starvation-free playing back, and achieve fairness among clients, we try to formulate the joint optimization of resource allocation and rate adaptation (JRA2) problem in this paper based on the information of streams' playout buffers and available network bandwidth. We first analyze the buffer behavior of a DASH stream using a modified version of M/D/1 queue and formulate the JRA2problem based on the analysis. To solve this problem, we propose an algorithm based on Generalize Benders Decomposition (GBD) to get optimal solution (JRA2-G) and devise one heuristic algorithm (JRA2-A) for acceleration. Based on the sufficient demonstrating results, the proposed algorithms can supply high-quality and smooth playing back for clients.

This paper addresses the problem of scheduling transmission requests in vehicular networks so that long-term road side unit (RSU) energy costs are minimized. We demonstrate that knowledge of vehicular routes greatly improves the energy service costs and request drop ratio of RSU transmission. At the same time, simple and fast prediction algorithms can recover a significant portion of the loss incurred by a lack of vehicle route knowledge. The proposed algorithms use recent historical traffic data and simple calculations, such as Bayesian estimates, to predict the next few routing decisions by a vehicle, in order to load balance the scheduling of its requests over the RSU network. Our simulation results show that, while the common assumption in the literature of knowing the vehicle routes is indeed crucial for achieving good performance, simple algorithms can be used in cases where vehicle routes are not known ahead of time, in order to achieve comparable costs and loss ratios.

Monitoring the dynamics of surface water using remote sensing technology is an essential research topic in many research areas. Change vector analysis (CVA) is a change detection method by constructing a vector of change based on multi-temporal images. It has been widely applied in land use and land cover change detection. Considering the specialty of surface water's spectral characteristics, it is anticipated that the CVA method can be revised and improved specially for monitoring surface water change. This study utilizes three Landsat images at different time phases for testing this proposed method, and uses traditional classification results for evaluation. Evaluation results demonstrate that surface water dynamic has been detected successfully using the CVA method. The proposed method along with its calibrated criteria can be applied for detecting surface water change between any other two time phases for this area. This can therefore be helpful for quick monitoring of water dynamic.

Microwave remote sensing has been largely applied to retrieve soil moisture (SM) from active and passive sensors. The obvious advantage of microwave sensors is that SM can be obtained regardless of atmospheric conditions. However, existing global SM products only provide observations at coarse spatial resolutions, which often hamper their application in regional hydrological studies. On the other hand, the vegetation temperature condition index (VTCI) has been widely used to monitor the SM status. It is based on high-spatial-resolution visible and infrared satellite observations. The aim of this study is to develop a simple and efficient downscaling approach for estimating accurate SM at higher spatial resolution. The VTCI calculated from the Moderate Resolution Imaging Spectroradiometer is used to downscale the coarse-resolution SM product that has been developed under the framework of the European Space Agency's Climate Change Initiative (CCI) projects. The original and downscaled SM estimates are further validated against thein situSM observations collected in the Yunnan province (southwest China). It is found that the accuracy level of CCI SM is similar to the results from previously published validation studies. The downscaled SM can maintain the accuracy of CCI SM and, at the same time, present more spatial details, demonstrating the feasibility of the proposed method. Overall, the notable advantages of the proposed method are simplicity, limited data requirements and purely relying on satellite measurements, and comparable accuracy level to other complex downscaling schemes. It will facilitate local hydrological applications, particularly in data-scarce regions, where the above-listed characteristics are important and useful.

Serious soil erosion is both the cause and the result of eco-environment degradation in Zuli River Basin (ZRB) of Chinese loess Plateau. The spatial distribution of soil erosion of ZRB in one humid year 1999 and dry year 2001 were estimated by modified SEMMED model which forced by the observed precipitation data, Digital Elevation Model, land use data derived from remote sensing data, and other field survey data under GIS environment. The results of simulated soil erosion suggest that the most serious soil erosion occurred in areas around Gejiacha, Caotan, Gejiazai and Jingyuan. The soil erosion in central and southern part of ZRB is less than that of Northern part. The Quwu Mountain, Yueliang Mountain and Huajialing have high vegetation cover and less soil erosion. These results indicate that the soil erosion is highly sensitive to the great interannual variability of rainfall intensity in arid regions such as ZRB.

Automated glacier mapping from satellite multispectral image data is very difficult because of the similar spectral characters of lateral and terminal moraine with that of neighbor rock. Paul (2002) has developed a method which integrates an ASTER multispectral image with a digital elevation model to map two debris-covered glaciers in the Alps. However, there are some problems when using the method in many mountain glaciers in the West of China. The critical slopes which identify the debris-covered tongue from rocks were tested first, which reveals that there is not a unique critical slope for all glaciers. Vectors from the China Glacier Inventory were used to derive expansion and shrinkage acting as a mask. A Temperature Index image was calculated by the one window algorithm from Landsat band 6 and NDVI. The method combines the advantage of Paul's method with the mask and Temperature Index. Most of the processing can be done automatically. It was applied to Keqikaer debris-covered glacier, west Tianshan Mountain, China.

This study analyzes long-term climate and glacier runoff changes to examine climate change and glacier mass balance response over the past 47 years by applied a modified monthly degree-day models in Tailan basin, the Tianshan Mountains of China. The results show that the snow line altitude is much closed to that of observation or other's result, which suggested the model, is valid to evaluate the glacier runoff changes. The mass balance of Tailan glaciers is all below zero after 1982-1983, the total mass balance is -16.07m during 1961-2007. The average discharge of glacier during 2001-2007 has increased 25% than that between 1961 to 1990, about 2% contributed from the increased of precipitation, and another 23% contributed from the lost of glacier mass balance. The 0.4 degree increase of summer air temperature has made -235.7mm yearly mass balance compensate with 18.3mm increase of annual precipitation, which equal to -635mm per degree mass balance without increase of precipitation. The result reveals that the contribution of glacier runoff to Tailan hydrological station has remarkable increase after 1990, especially after 2000.

The source area of the first phase of the west route scheme of South to North Water Diversion Project in China (SNWD) is located in the southeast part of the Tibetan Plateau, where the altitude is high with fragile eco-environment, and is lack of observed hydrology data. One of the important parts of SNWD is estimating the monthly water discharge of each dam in last 40 years, which only had less than 4 years observed hydrological data. The monthly discharge in last 40 years only were calculated by other downstream stations, the basis of the calculation is that they have similar hydrology character, so the aim of this research is compare physiographic characters of 6 sub-basins in source region via Remote Sensing, GIS and observed data. The study area lies in the range of 98°45'-102°50'E and 30°45'-33°50'N, coving an area of 49159 km2. Field observations with GPS were made to identify different land cover's feather signature dataset, and some soil samples were collected to analysis the physical feathers, such as particle size and the content of organic carbon. Sixty images of 1:100000 topographic maps were mosaic and five Landsat Themetic Mapper (TM\\ETM+) images were used for visual image interpretation. 1: 250000 scales DEM were used for calculating the statistic, such as mean, maximum, minimum, variance, middle, of elevation, slope, aspect and topographic index (CTI), and yearly precipitation and yearly average air temperature. Land cover were compared too. According to the calculating and interpreting, the hydrological characters or the natural environment of each watershed were analyzed, whole area has more similar, but some difference exists. The Niqu reach has the highest basin elevation average, the least basin slope mean, and the largest average CTI and the least areal runoff, and more swamp, which should have the least runoff modulus. The Andou sub-basin has the lower basin elevation average, the highest basin slope mean, and the lowest average CTI and the biggest areal runoff, which should have the biggest runoff modulus. The watershed above Zhuba has the very similar hydrological character with the merged area of 6 subbasins, and may has the same runoff modulus. © 2005 IEEE.

It is very important to study the hydrological process and the potential change of water resource under future climate scene either in science or in practice. However, no further research on the hydrological process in cold regions due to the high altitude, poor natural environment and rare observing sites at the present time. The upper reach of Yellow River is at northeast part of Tibet Plateau, located between 32degN-36deg25'N and 95deg30'E-103deg30'E, and covers approximately 120,000 km2, the average altitude above sea level of which is about 3500 meters. In this study, a macro-scale hydrology model, VIC-3L, via Remote Sensing and GIS, was used to tested and verified by using Tangnaihai and Huangheyan hydrological stations' observed date, then simulated the potential consequences of 2010s, 2030s and 2050s. It is proved that the degradation of permafrost in 2050s had the greatest influence in discharge, the increase of evaporation, which caused by the increase of temperature and the degradation of permafrost, would exceed the increase of precipitation. The discharge reaches the maximum in 2010s under the decrease of temperature 0.1 and the increase of precipitation 22%