Neighbor Weight Research Articles

A Large-Scale Evaluation Of Shape-Aware Neighborhood Weights And Neighborhood Sizes

In this paper, we define and evaluate a weighting scheme for neighborhoods in point sets. Our weighting takes the shape of the geometry, i.e., the normal information, into account. This causes the obtained neighborhoods to be more reliable in the sense that connectivity also depends on the orientation of the point set. We utilize a sigmoid to define the weights based on the normal variation. For an evaluation of the weighting scheme, we turn to a Shannon entropy model for feature classification that can be proven to be non-degenerate for our family of weights. Based on this model, we evaluate our weighting terms on a large scale of both clean and real-world models. This evaluation provides results regarding the choice of optimal parameters within our weighting scheme. Furthermore, the large-scale evaluation also reveals that neighborhood sizes should not be fixed globally when processing models. Finally, we highlight the applicability of our weighting scheme within the application context of denoising.

ROAD TYPE CLASSIFICATION OF MLS POINT CLOUDS USING DEEP LEARNING

Abstract. Functional classification of the road is important to the construction of sustainable transport systems and proper design of facilities. Mobile laser scanning (MLS) point clouds provide accurate and dense 3D measurements of road scenes, while their massive data volume and lack of structure also bring difficulties in processing. 3D point cloud understanding through deep neural networks achieves breakthroughs since PointNet and arouses wide attention in recent years. In this paper, we study the automatic road type classification of MLS point clouds by employing a point-wise neural network, RandLA-Net, which is designed for consuming large-scale point clouds. An effective local feature aggregation (LFA) module in RandLA-Net preserves the local geometry in point clouds by formulating an enhanced geometric feature vector and learning different point weights in a local neighborhood. Based on this method, we also investigate possible feature combinations to calculate neighboring weights. We train on a colorized point cloud from the city of Hannover, Germany, and classify road points into 7 classes that reveal detailed functions, i.e., sidewalk, cycling path, rail track, parking area, motorway, green area, and island without traffic. Also, three feature combinations inside the LFA module are examined, including the geometric feature vector only, the geometric feature vector combined with additional features (e.g., color), and the geometric feature vector combined with local differences of additional features. We achieve the best overall accuracy (86.23%) and mean IoU (69.41%) by adopting the second and third combinations respectively, with additional features including Red, Green, Blue, and intensity. The evaluation results demonstrate the effectiveness of our method, but we also observe that different road types benefit the most from different feature settings.

Weak consistency for the estimators in a semiparametric regression model based on negatively associated random errors

In this article, we consider the semiparametric regression model: , which is an important and very useful statistical model. We investigate the parametric component and nonparametric component estimators in a semiparametric regression model based on negatively associated random errors. The weak consistency for the estimators and of and respectively are established under some suitable conditions. In addition, a simulation to study the numerical performance of the consistency for the nearest neighbor weight function estimators is provided and a real data application is then presented.

Improvement of Grey System Model using Particle Swarm Optimization

An improvement of the traditional grey system model, GM(1,1), to enhance forecast accuracy, has been realized using the particle swarm optimization (PSO) algorithm. Unlike the GM(1,1) which uses a fixed adjacent neighbor weight for all data sets, the proposed PSO-improved model, PSO-GM(1,1), determines an optimal adjacent neighbor weight, based on the presented data set. This optimal adjacent neighbor weight so determined is the principal factor that enhances forecast accuracy. The performance of the proposed model was evaluated using generated monotonic increasing and decreasing data sets as well as measured energy consumption data for a laptop computer, desktop computer, printer, and photocopier. The performance of PSO-GM(1,1) was compared with that of GM(1,1), and two other models in literature that sought to improve the performance of GM(1,1). The PSO-GM(1,1) outperformed the traditional model and the two other models. For the monotonic increasing data, the mean absolute percentage error (MAPE) for the proposed model was 0.007% as against a MAPE value of 20.383% for the GM(1,1). For the monotonic decreasing data, the PSO-GM(1,1) again outperformed GM(1,1), yielding a MAPE of 0.057% compared to a value of 13.407% for the traditional model. For the measured laptop computer energy data, the obtained MAPE for the PSO-GM(1,1) was 0.675% while the values for the two models were 4.052% and 2.991%. For the measured desktop computer energy data, the obtained MAPE for the PSO-GM(1,1) was 0.0018% while the values for the two models were 0.0018% and 1.163%. For the data associated with the printer, the MAPEs were 8.414% for the PSO-GM(1,1), 20.957% for the first model and 9.080% for the second model. For the measured photocopier energy data, the obtained MAPE for the PSO-GM(1,1) was 0.901% while the values for the two models were 3.799% and 0.943%. Thus, the proposed PSO-GM(1,1) greatly improves forecast accuracy and is recommended for adoption, for forecasting.

Anomaly Based Sea-Surface Small Target Detection Using K-Nearest Neighbor Classification

Sea-surface small target detection is always a difficult problem in high-resolution maritime ubiquitous radars for complex characteristics of sea clutter, weak target returns, and diversity of targets. Multiple features extracted from radar returns in different domains have ability but not enough to solely distinguish radar returns with target from sea clutter. Joint exploitation of multiple features becomes the key to improve detection performance. In this article, the K-nearest neighbor (KNN) algorithm and anomaly detection idea are cooperated to develop a novel sea-surface target detection method in the feature space spanned by the eight existing salient features. The detection is realized by the anomaly detection followed by a specially designed KNN-based classifier with a controllable false alarm rate. In the anomaly detection, a decision region is determined by the hyper-spherical coverage of the training set of sea clutter that is sufficient and ergodic in the feature space. The KNN-based classifier is designed based on the training sample set of sea clutter and the training sample set of simulated target returns plus sea clutter that is sufficient but nonergodic, by joint usage of feature weighting, neighbor weighting, and distance weighting. The novel method is validated by the two open and recognized IPIX and CSIR radar databases for sea-surface small target detection. The results show that it provides significant performance improvement in comparison with the existing multiple-feature-based detection methods, owing to the fact that the novel method avoids the dimension restriction and feature compression loss in the existing methods.

D-TSVR Recurrence Prediction Driven by Medical Big Data in Cancer

Secondary use of medical big data is becoming increasingly popular in healthcare services and clinical research in the medical industry. Cancer recurrence is a common phenomenon of cancer patients after treatment (recovery period). Studying the time and influencing factors of cancer recurrence can provide effective clinical intervention means, which is the gospel of cancer patients. In this article, a sample of 50 000 cases of seven cancer patients, including liver cancer, lung cancer, kidney cancer, breast cancer, uterine cancer, stomach cancer, and bowel cancer, was collected. A twin support regression vector machine (TSVR) algorithm based on dependent nearest neighbor (DNN) weighting is proposed, the eplion-TSVR model is improved by DNN-weighted algorithm with local information mining function, and the solution of the improved model is derived. It is proposed to use the cuckoo algorithm to determine the optimal parameters of DNN to determine the optimal dependency region domain. In this article, the improved TSVR algorithm is used to establish a cancer recurrence prediction model. The prediction accuracy of the model for various cancers can reach more than 91%, which is significantly higher than that of convolutional neural network and e-TSVR models.

A Fast Radio Map Construction Method Merging Self-Adaptive Local Linear Embedding (LLE) and Graph-Based Label Propagation in WLAN Fingerprint Localization Systems.

Indoor WLAN fingerprint localization systems have been widely applied due to the simplicity of implementation on various mobile devices, including smartphones. However, collecting received signal strength indication (RSSI) samples for the fingerprint database, named a radio map, is significantly labor-intensive and time-consuming. To solve the problem, this paper proposes a semi-supervised self-adaptive local linear embedding algorithm to build the radio map. First, this method uses the self-adaptive local linear embedding (SLLE) algorithm based on manifold learning to reduce the dimension of the high-dimensional RSSI samples and to extract a neighbor weight matrix. Secondly, a graph-based label propagation (GLP) algorithm is employed to build the radio map by semi-supervised learning from a large number of unlabeled RSSI samples to a few labeled RSSI samples. Finally, we propose a k self-adaptive neighbor weight (kSNW) algorithm, used for radio map construction in this paper, to realize online localization. The results of the experiments conducted in a real indoor environment show that the proposed method reduces the demand for large quantities of labeled samples and achieves good positioning accuracy. With only 25% labeled RSSI samples, our system can obtain positioning accuracy of more than 88%, within 3 m of localization errors.

Enhanced Graph Isomorphism Network for Molecular ADMET Properties Prediction

The evaluation of absorption, distribution, metabolism, exclusion, and toxicity (ADMET) properties plays a key role in a variety of domains including industrial chemicals, agrochemicals, cosmetics, environmental science, food chemistry, and particularly drug development. Since molecules are often intrinsically described as molecular graphs, graph neural networks have recently been studied to improve the prediction of ADMET properties. Among many graph neural networks published in recent years, Graph Isomorphism Network (GIN) is a relatively recent and very promising one. In this paper, we propose an enhanced GIN, called MolGIN, via exploiting the bond features and differences influence of the atom neighbors to end-to-end predict ADMET properties. Based on GIN, MolGIN concatenates the bond feature together with node feature in the feature aggregator and applies a gate unit to adjust the atomic neighborhood weights to map the differences in the interaction strength between the central atom and its neighbors, such that more meaningful structural patterns of molecules can be explored toward better molecular modeling. Extensive experiments were conducted on seven public datasets to evaluate MolGIN against four baseline models with benchmark metrics. Experimental results of MolGIN were also compared with state-of-the-art results published in the last three years on each dataset. Experimental results in terms of RMSE and AUC show that MolGIN significantly boosts the prediction performance of GIN and markedly outperforms the baseline models, and achieves comparable or superior performance to state-of-the-art results.

On Signed Domination of Grid Graph

Let G(V, E) be a finite connected simple graph with vertex set V(G). A function is a signed dominating function f : V(G)→{−1,1} if for every vertex v ∈ V(G), the sum of closed neighborhood weights of v is greater or equal to 1. The signed domination number γs(G) of G is the minimum weight of a signed dominating function on G. In this paper, we calculate the signed domination numbers of the Cartesian product of two paths Pm and Pn for m = 6, 7 and arbitrary n.

A Multi-objective Improved Squirrel Search Algorithm based on Decomposition with External Population and Adaptive Weight Vectors Adjustment

In order to further improve the performance of the Multi-objective Evolutionary Algorithm Based on Decomposition (MOEA/D) in solving multi-objective optimization problems, this paper constructs a multi-objective optimization algorithm by taking MOEA/D as the multi-objective framework and the Squirrel Search Algorithm (SSA) as the core evolutionary strategy. Besides, both of them are modified and the Multi-objective Improved Squirrel Search Algorithm based on Decomposition with External Population and Adaptive Weight Vectors Adjustment (MOEA/D-EWA-ISSA) is proposed. MOEA/D-EWA-ISSA establishes an external population for every individual to retain the evolutionary information and maintain the population diversity, external individuals participate in evolution and produce better offspring, the convergence and distribution of Pareto Front (PF) are improved. As for SSA, the jumping search method and the progressive search method are introduced to it, different evolutionary strategies are provided to solve subproblems, which further improves the ability of core evolutionary strategy to solve subproblems and the convergence of the obtained PF. Furthermore, MOEA/D-EWA-ISSA adjusts every weight vector adaptively according to the population’s actual evolutionary direction and the representative neighbor weight vectors, the distribution of the obtained PF is improved as well. The experimental results on multi-objective test functions show that the convergence and distribution of PF have obvious improvement when the improved SSA, the improved MOEA/D and the whole MOEA/D-EWA-ISSA are used to solve multi-objective optimization problems.

Complete q-th moment convergence and its statistical applications

In this paper, we mainly investigate the complete q-th moment convergence for sums of widely orthant dependent random variables by utilizing the Rosenthal type moment inequality and truncation method. Some sufficient conditions for the complete q-th moment convergence are provided. The results obtained in the paper generalize some corresponding ones of negatively associated random variables in the literature. As an application, the complete consistency for the weighted estimator in a nonparametric regression model is established, and the simulation studies are presented to show the consistency for the nearest neighbor weight function estimator in a nonparametric regression model.

Application of improved Multiscale normalization

In this paper, a method based on neighborhood information is proposed, which adjusts the parameters by using a method of chi square with Gauss weighted distance and the self-adaptability of shared neighbor weighting. And by integrating the nearest neighbor weighted adaptive method, each pixel is automatically given a scale parameter to reduce the need to adjust the parameters. Gauss weighted local neighborhood information is introduced in this paper, to construct the similarity matrix directly on the original image. At the same time, based on the existing graph-cut image segmentation method is necessary to construct a better network graph by using the maximum flow minimum cut. So, a method is applicated, NJW algorithm is replaced by a improved Graph cut which combines the multi-scale analysis method, to improve the efficiency of the original segmentation method and enhance the clinical usability.

A note on the asymptotic properties of the estimators in a semiparametric regression model

In this paper, we investigate the parametric component and nonparametric component estimators in a semiparametric regression model based on -mixing random errors. The r-th mean consistency and uniform consistency are established under some suitable conditions. Finally, a simulation to study the numerical performance of the consistency for the nearest neighbor weight function estimators is provided. The results obtained in the paper complement the existing ones of independent random errors to the case of -mixing random errors.

Modeling and Optimizing of the Multi-Layer Nearest Neighbor Network for Face Image Super-Resolution

In this paper, we propose a face super-resolution (FSR) method to handle the decreasing face recognition rate caused by low-quality images. To better model the input images, we build a nearest neighbor network (NNN) which consists of nodes and paths by introducing the second-layer nearest neighbors (SLNNs), where the paths of the network represent the distance between nodes. As the SLNN is trained in the high-resolution (HR) space and is exponentially supplementary to the traditional first-layer nearest neighbors (FLNNs), the neighbor inadequacy problem can be effectively solved by enriching the neighbor candidate set via NNN. Furthermore, we solve the NNN for the optimal weights of neighbors. Finally, we fuse the refined weights and neighbors for better reconstruction results. The effectiveness of this fusion strategy is validated by both quantitative and qualitative experimental results. The extensive experimental results on the public face datasets and real-world challenging low-resolution (LR) images demonstrate that the proposed method performs favorably against the state-of-the-art methods.

A Spectral Unmixing Method by Maximum Margin Criterion and Derivative Weights to Address Spectral Variability in Hyperspectral Imagery

Limited to the low spatial resolution of the hyperspectral imaging sensor, mixed pixels are inevitable in hyperspectral images. Therefore, to obtain the endmembers and corresponding fractions in mixed pixels, hyperspectral unmixing becomes a hot spot in the field of remote sensing. Endmember spectral variability (ESV), which is common in hyperspectral images, affects spectral unmixing accuracy. This paper proposes a spectral unmixing method based on maximum margin criterion and derivative weights (MDWSU) to reduce the effect of ESV on spectral unmixing. Firstly, in the MDWSU model, an effective and fast algorithm is employed for establishing the endmember spectral library. Then a spectral weighting matrix based on the maximum margin criterion is constructed based on the endmember spectral library. Besides, derivative analysis and local neighborhood weights are merged into local neighborhood derivative weights, which act as a regularization term to penalize different abundance vectors. Local neighborhood derivative weights and spectral weighting matrix are proved to reduce the effect of ESV. Real hyperspectral data experiments show that the MDWSU model can obtain more accurate endmembers and abundance estimation. In addition, the experimental results, including the spectral angle distance and the root mean square error, prove the superiority of the MDWSU model over the previous methods.

ABNE: An Attention-Based Network Embedding for User Alignment Across Social Networks

User alignment across social networks can facilitate more information/knowledge transferring across networks and thereby benefit several applications, including social link prediction, cross-domain recommendation, and information diffusion. Several works try to learn a common subspace for networks by preserving the structural proximities, such that different contribution weights of neighbors are ignored as users were always connected by unweighted edges. In this paper, we propose an attention-based network embedding model that exploits the social structures for user alignment. In particular, two main components are contained in our model framework: a masked graph attention mechanism which tries to learn the alignment task driven attention weights by the supervision of pre-aligned user pairs, and an embedding algorithm tries to learn a common vector space by explicitly modeling the weighted contribution probabilities between follower-ships and followee-ships. With the learned weights and embeddings transferring between these two components, we construct a unified model for user embedding and alignment. Stochastic gradient descent and negative sampling are adopted for efficient learning and scalability. The extensive experiments on real-world social network data sets demonstrate the effectiveness and robustness of the proposed model compared with several state-of-the-art methods.

A robust adaptive spatial and temporal image fusion model for complex land surface changes

Spatial and temporal satellite image fusion (STIF) has provided a feasible alternative for generating imagery with both high spatial and temporal resolution, thus expanding the applications of existing satellite sensors. However, a critical challenge confronting the further development of STIF is to systematically and robustly address complex land surface changes, which include land cover changes without shape changes (e.g., crop rotation) and land cover changes with shape changes (e.g., urban expansion), in addition to conventional land surface changes (e.g., phenological changes of vegetation). This paper presents the Robust Adaptive Spatial and Temporal Fusion Model (RASTFM) to tackle this challenge with one prior pair of MODIS-Landsat images. In RASTFM, land surface changes are reorganized into non-shape changes (including phenological changes and land cover changes without shape changes) and shape changes (i.e., land cover changes with shape changes), which are handled differently. However, both non-shape changes and shape changes are predicted through a Non-Local Linear Regression (NL-LR) of the subject pixel's similar neighbors. A regression based high-pass modulation is also performed as a post-processing step to improve both the spatial details and spectral fidelity of the predicted Landsat image. Unlike other STIF models (e.g., the Spatial and Temporal Adaptive Reflectance Fusion Model, STARFM), RASTFM can find similar neighboring pixels more precisely through a non-local searching strategy and derives the weights of similar neighbors more rigorously via a linear regression model. As both non-shape and shape changes are treated based on the regression of similar neighboring pixels, the land surface changes are processed in a unified manner. Experiments that use one simulated and three actual MODIS-Landsat datasets featured by different types of land surface changes were conducted to demonstrate the performance of RASTFM. Comparisons with the state-of-the-art STIF models, including weighted function, unmixing and dictionary-learning methods, show that NL-LR based RASTFM can capture the land surface changes in various landscapes more accurately and robustly in a unified manner, thereby facilitating the continuous and detailed monitoring of complex and diverse land surface dynamics.

Complete moment convergence for negatively orthant dependent random variables and its applications in statistical models

In this paper, a general result on complete moment convergence for arrays of rowwise negatively orthant dependent random variables is obtained. In addition, we present some sufficient conditions to prove the complete moment and complete convergences for the variables. As applications, the complete consistency for the estimators of nonparametric and semiparametric regression models based on negatively orthant dependent errors is established by using the complete convergence that we established. A simulation to study the numerical performance of the consistency for the nearest neighbor weight function estimator in semiparametric regression model is given. Our results generalize and improve some corresponding ones for independent random variables and negatively associated random variables.

Detecting epileptic seizure with different feature extracting strategies using robust machine learning classification techniques by applying advance parameter optimization approach.

Epilepsy is a neurological disorder produced due to abnormal excitability of neurons in the brain. The research reveals that brain activity is monitored through electroencephalogram (EEG) of patients suffered from seizure to detect the epileptic seizure. The performance of EEG detection based epilepsy require feature extracting strategies. In this research, we have extracted varying features extracting strategies based on time and frequency domain characteristics, nonlinear, wavelet based entropy and few statistical features. A deeper study was undertaken using novel machine learning classifiers by considering multiple factors. The support vector machine kernels are evaluated based on multiclass kernel and box constraint level. Likewise, for K-nearest neighbors (KNN), we computed the different distance metrics, Neighbor weights and Neighbors. Similarly, the decision trees we tuned the paramours based on maximum splits and split criteria and ensemble classifiers are evaluated based on different ensemble methods and learning rate. For training/testing tenfold Cross validation was employed and performance was evaluated in form of TPR, NPR, PPV, accuracy and AUC. In this research, a deeper analysis approach was performed using diverse features extracting strategies using robust machine learning classifiers with more advanced optimal options. Support Vector Machine linear kernel and KNN with City block distance metric give the overall highest accuracy of 99.5% which was higher than using the default parameters for these classifiers. Moreover, highest separation (AUC=0.9991, 0.9990) were obtained at different kernel scales using SVM. Additionally, the K-nearest neighbors with inverse squared distance weight give higher performance at different Neighbors. Moreover, to distinguish the postictal heart rate oscillations from epileptic ictal subjects, and highest performance of 100% was obtained using different machine learning classifiers.

SOM-Based Class Discovery for Emotion Detection Based on DEAP Dataset

This paper investigates the possibility of identifying classes by clustering. This study includes employing Self-Organizing Maps (SOM) in identifying clusters from EEG signals that could then be mapped to emotional classes. Beginning by training varying sizes of SOM with the EEG data provided from the public dataset: DEAP. The produced graphs showing Neighbor Distance, Sample Hits, and Weight Position are examined. Following that, the ground-truth label provided in DEAP is tested, in order to identify correlations between the label and the clusters produced by the SOM. The results show that there is a potential of class discovery using SOM-based clustering. It is then concluded that by evaluating the implications of this work and the difficulties in evaluating its outcome.