Similar Neighbors Research Articles

Do poverty and wealth look the same the world over? A comparative study of 12 cities from five high-income countries using street images

Urbanization and inequalities are two of the major policy themes of our time, intersecting in large cities where social and economic inequalities are particularly pronounced. Large scale street-level images are a source of city-wide visual information and allow for comparative analyses of multiple cities. Computer vision methods based on deep learning applied to street images have been shown to successfully measure inequalities in socioeconomic and environmental features, yet existing work has been within specific geographies and have not looked at how visual environments compare across different cities and countries. In this study, we aim to apply existing methods to understand whether, and to what extent, poor and wealthy groups live in visually similar neighborhoods across cities and countries. We present novel insights on similarity of neighborhoods using street-level images and deep learning methods. We analyzed 7.2 million images from 12 cities in five high-income countries, home to more than 85 million people: Auckland (New Zealand), Sydney (Australia), Toronto and Vancouver (Canada), Atlanta, Boston, Chicago, Los Angeles, New York, San Francisco, and Washington D.C. (United States of America), and London (United Kingdom). Visual features associated with neighborhood disadvantage are more distinct and unique to each city than those associated with affluence. For example, from what is visible from street images, high density poor neighborhoods located near the city center (e.g., in London) are visually distinct from poor suburban neighborhoods characterized by lower density and lower accessibility (e.g., in Atlanta). This suggests that differences between two cities is also driven by historical factors, policies, and local geography. Our results also have implications for image-based measures of inequality in cities especially when trained on data from cities that are visually distinct from target cities. We showed that these are more prone to errors for disadvantaged areas especially when transferring across cities, suggesting more attention needs to be paid to improving methods for capturing heterogeneity in poor environment across cities around the world.

Phonological conditioning of affricate variability in Emirati Arabic

This study investigates the conditioning effects of neighbouring consonants on the realisation of the phonemes /k/ and /dʒ/ in Emirati Arabic (EA), which are optionally realised as [tʃ] and [j], respectively. Based on previous accounts of EA and other Gulf Arabic (GA) dialects, we set out to test the prediction that proximity of other, phonetically similar coronal (COR) obstruents [COR, −son, −cont] and coronal postalveolar fricatives [COR, −ant] inhibit the surface realisation of the affricate variants of these phonemes. We examine elicitation data from twenty young female native speakers of EA, using stimuli with the target segment in the presence of a similar neighbour, as compared to words with the neighbour at a longer distance or with another coronal consonant. The results point to an asymmetry in the behaviour of the voiced and voiceless targets, such that the predicted inhibitory effect is confirmed for the voiced, but not the voiceless target. We argue that this finding, coupled with a consideration of the intra-participant and lexical trends in the data, is compatible with an approach that treats the two processes as being at different stages of development, where the [k∼tʃ] alternation is a completed phonemic change, while the [dʒ∼j] alternation is a synchronic phonological process.

Fraud detection on multi-relation graphs via imbalanced and interactive learning

In recent years, the application of Graph Neural Networks (GNNs) in fraud detection has gained considerable attention. GNNs have demonstrated their efficacy in leveraging the abundant relational information inherent in graph-structured data for such tasks. However, despite the remarkable progress achieved thus far, several challenges still need to be addressed in the current GNN-based algorithms. On the one hand, GNNs exhibit limited performance when confronted with imbalanced label distribution between fraudsters and benign users. On the other hand, the embedding representations learned for nodes and their neighbors in the last layer are frequently processed in a simplistic manner, such as concatenation or averaging, which may compromise the performance of downstream tasks. To alleviate the above problems, we propose a novel imbalanced and interactive learning framework for fraud detection on multi-relation graphs (IMINF for short). Concretely, firstly, we design a novel neighbor sampler that can be trained using supervised contrastive learning. This sampler selects the most similar and consistent neighbors to update the target node's features, taking into account label information from multiple perspectives. Secondly, we introduce the learnable contrastive embedding and an improved relational aggregator to enhance the representation of the target node. Thirdly, we propose a new interactive learning module, which consists of explicit and implicit transformation layers, with the aim of capturing deep interactive relations between a target node and its neighbors. Additionally, we devise a feature mapping module prior to the relation aggregation step, transferring the raw input features to a latent feature subspace. Extensive experiments on two real-world fraud detection datasets demonstrate that the proposed model outperforms existing baselines across all evaluation metrics, especially for the Amazon fraud detection dataset.

Local nonlinear dimensionality reduction via preserving the geometric structure of data

Dimensionality reduction has many applications in data visualization and machine learning. Existing methods can be classified into global ones and local ones. The global methods usually learn the linear relationship in data, while the local ones learn the manifold intrinsic geometry structure, which has a significant impact on pattern recognition. However, most of existing local methods obtain an embedding with eigenvalue or singular value decomposition, where the computational complexities are very high in a large amount of high-dimensional data. In this paper, we propose a local nonlinear dimensionality reduction method named Vec2vec, which employs a neural network with only one hidden layer to reduce the computational complexity. We first build a neighborhood similarity graph from the input matrix, and then define the context of data points with the random walk properties in the graph. Finally, we train the neural network with the context of data points to learn the embedding of the matrix. We conduct extensive experiments of data classification and clustering on nine image and text datasets to evaluate the performance of our method. Experimental results show that Vec2vec is better than several state-of-the-art dimensionality reduction methods, except that it is equivalent to UMAP on data clustering tasks in the statistical hypothesis tests, but Vec2vec needs less computational time than UMAP in high-dimensional data. Furthermore, we propose a more lightweight method named Approximate Vec2vec (AVec2vec) with little performance degradation, which employs an approximate method to build the neighborhood similarity graph. AVec2vec is still better than some state-of-the-art local dimensionality reduction methods and competitive with UMAP on data classification and clustering tasks in the statistical hypothesis tests.

NetPro: Neighborhood Interaction-based Drug Repositioning via Label Propagation.

Drug repositioning is an important approach for predicting new disease indications of the existing drugs in drug discovery. A great progress has been achieved in drug repositioning. However, effectively utilizing the localized neighborhood interaction features of drug and disease in drug-disease associations remains challenging. This paper proposes a neighborhood interaction-based method called NetPro for drug repositioning via label propagation. In NetPro, we first formulate the known drug-disease associations, various disease and drug similarities from different perspectives to construct drug-drug and disease-disease networks. Meanwhile we employ the nearest neighbors and their interactions in the constructed networks to devise a new approach for computing drug similarity and disease similarity. To implement the prediction of new drugs or diseases, a preprocessing step is applied to renew the known drug-disease associations using our calculated drug and disease similarities. We then employ a label propagation model to predict drug-disease associations by the drug and disease linear neighborhood similarities derived from the renewed drug-disease associations. The experimental results on three benchmark datasets show that NetPro can effectively identify potential drug-disease associations and achieve better prediction performance than the existing methods. Case studies further demonstrate that NetPro is capable of predicting promising candidate disease indications for drugs.

Deep Cross-Network Alignment with Anchor Node Pair Diverse Local Structure

Network alignment (NA) offers a comprehensive way to build associations between different networks by identifying shared nodes. While the majority of current NA methods rely on the topological consistency assumption, which posits that shared nodes across different networks typically have similar local structures or neighbors, we argue that anchor nodes, which play a pivotal role in NA, face a more challenging scenario that is often overlooked. In this paper, we conduct extensive statistical analysis across networks to investigate the connection status of labeled anchor node pairs and categorize them into four situations. Based on our analysis, we propose an end-to-end network alignment framework that uses node representations as a distribution rather than a point vector to better handle the structural diversity of networks. To mitigate the influence of specific nodes, we introduce a mask mechanism during the representation learning process. In addition, we utilize meta-learning to generalize the learned information on labeled anchor node pairs to other node pairs. Finally, we perform comprehensive experiments on both real-world and synthetic datasets to confirm the efficacy of our proposed method. The experimental results demonstrate that the proposed model outperforms the state-of-the-art methods significantly.

ISRToken: Learning similarities tokens for precise infrared spectrum recognition model via transformer

Infrared spectrum recognition (ISR) has been widely used in the fields of biomedical imaging and cancer diagnosis. However, existing methods cannot deal with extreme infrared spectral random noise and band overlap. To address these challenges, we identify three cues from infrared spectrum, namely, neighborhood similarities, significant spectral structure changes, and key minority relationships. To leverage the observed findings, we propose a novel key minority relationship-aware method based on the Transformer architecture in which the spectral section relationships can be learned. Specifically, we design several orientation tokens to explicitly encode the basic similarities regions. Meanwhile, a novel token guide multi-loss function is designed to guide the similarities tokens as they learn the desired regional similarities and relationships. We evaluate the proposed method on three challenging benchmark ISR datasets. Experiments show that our method achieves better performance compared with state-of-the-art methods.

Research on seismic hydrocarbon prediction based on a self-attention semi-supervised model

Predictions of oil and gas reservoirs play an essential role in hydrocarbon reservoir exploration. Multi-attribute seismic data can provide abundant reservoir information. However, traditional reservoir prediction models only reflect the characteristics of local seismic channels, ignoring information on global geological spatial structures. Thus, a semi-supervised prediction model based on self-attention is proposed to extract reservoirs' oil and gas characteristics. With the high cost of labeled samples, neighborhood similarity measurement based on Grey Relation Analysis is applied to determine the homogeneity of adjacent seismic traces for the first time to obtain pseudo-labels while considering geological spatial structure. As a semi-supervised prediction model, a multilayer convolutional neural network with a multi-head self-attention mechanism is utilized, which can extract features from multiple dimensions and investigate global structural information. The experimental results demonstrate that the algorithm in this paper's reservoir oil and gas prediction is more accurate than that of conventional approaches and is consistent with real drilling data.

Video SAR Moving Target Shadow Detection Based on Intensity Information and Neighborhood Similarity

Video Synthetic Aperture Radar (SAR) has shown great potential in moving target detection and tracking. At present, most of the existing detection methods focus on the intensity information of the moving target shadow. According to the mechanism of shadow formation, some shadows of moving targets present low contrast, and their boundaries are blurred. Additionally, some objects with low reflectivity show similar features with them. These cause the performance of these methods to degrade. To solve this problem, this paper proposes a new moving target shadow detection method, which consists of background modeling and shadow detection based on intensity information and neighborhood similarity (BIIANS). Firstly, in order to improve the efficiency of image sequence generation, a fast method based on the Back-projection imaging algorithm (f-BP) is proposed. Secondly, due to the low-rank characteristics of stationary objects and the sparsity characteristics of moving target shadows presented in the image sequence, this paper introduces the low-rank sparse decomposition (LRSD) method to perform background modeling for obtaining better background (static objects) and foreground (moving targets) images. Because the shadows of moving targets appear in the same position in the original and the corresponding foreground images, the similarity between them is high and independent of their intensity. Therefore, using the BIIANS method can obtain better shadow detection results. Real W-band data are used to verify the proposed method. The experimental results reveal that the proposed method performs better than the classical methods in suppressing false alarms, missing alarms, and improving integrity.

An automatic band selection algorithm based on connection centre evolution

ABSTRACT The characteristics of high dimensionality, strong band correlation, and high information redundancy make band selection become a necessary step for efficient processing of hyperspectral images. In this paper, a new band selection method based on connection centre evolution (CCE) is proposed, named -nearest neighbour CCE band selection (NNCCEBS). First, according to CCE, we measure the band similarity under different observation scales by iteratively calculating the connectivity at different scales. Therefore, NNCCEBS can achieve multi-scale band selection, and is able to complete it at one time. Next, to avoid the selected bands being aggregated, an -nearest neighbour similarity matrix is introduced, which only retains the connectivity of -nearest bands for a band. By doing so, the new method is less sensitive to the parameter settings, and can select more accurate band centres. Experiments on real hyperspectral datasets prove the effectiveness of our method.

Firearm Legislation - The Association between Neighboring States and Crude Death Rates.

Few studies have examined the impact of interstate differences in firearm laws on state-level firearm mortality. We aim to study the association between neighboring states' firearm legislation and firearm-related crude death rate (CDR). The CDC Web-based Injury Statistics Query and Reporting System (WISQARS) was queried for adult all-intent (accidental, suicide, and homicide) firearm-related CDR among the 50 states from 2012 to 2020. States were divided into five cohorts based on the Giffords Law Center Annual Gun Law Scorecard and two groups were constructed: Strict (A, B, C) and Lenient (D, F). We examined the effect of 1) a single incongruent neighbor, defined as "Different" if the state is bordered by ≥1 state with a grade score difference > 1, and 2) the average grade of all neighboring states, defined as "Different" if the average of all neighboring states resulted in a grade score difference > 1. Strict states with similar average neighbors had significantly lower CDR compared to Strict states with different average neighbors (2.98 [1.91-5.06] vs. 3.87 [2.37-5.94], p = 0.02) while Lenient states with similar average neighbors had significantly higher CDR compared to Lenient states with different average neighbors (6.02 [4.56-8.11] vs. 4.7 [3.95-5.35], p = 0.002). Lenient states surrounded by all similar Lenient states had the highest CDR, which was significantly higher than Lenient states with ≥1 different neighbor (6.52 [5.09-8.96] vs. 5.19 [3.85-6.61], p < 0.001). However, Strict states with ≥1 different neighbor did not have higher CDR compared to Strict states surrounded by all similar Strict states (3.39 [2.17-5.35] vs. 3.14 [1.91-5.38], p = 0.5). We report a lopsided neighboring effect whereby Lenient states may benefit from at least 1 Strict neighbor while Strict states may be adversely affected only when surrounded by mostly Lenient neighbors. These findings may assist policymakers regarding the efficacy of their own state's legislation in the context of incongruent neighboring states. IV, retrospective observational.

Contrasting seed traits of co-existing seeds lead to a complex neighbor effect in a seed-rodent interaction.

Scatter-hoarding rodents play important roles in seed dispersal and predation in many forest ecosystems. Existing studies have shown that the seed foraging preference of rodents is directly affected by seed traits and indirectly affected by the traits of other co-existing seeds nearby (i.e., neighbor effect). Plant seeds exhibit a combination of diverse seed traits, including seed size, chemical defense, and nutrient content. Therefore, it is difficult to evaluate the influence of each single seed trait on such neighbor effects. Here, by using artificial seeds, we investigated the impacts of contrasts in seed size, tannin content, and nutrient content on neighbor effects. We tracked 9000 tagged artificial seeds from 30 seed-seed paired treatments in a subtropical forest in southwest China. The contrast in seed size between paired seeds created obvious neighbor effects measured through three seed dispersal related indicators: the proportion of seeds being removed, the proportion of seeds cached, and the distance transported by rodents. However, the magnitudes and the signs of the neighbor effects differed among pairs, including both apparent mutualism and apparent competition, depending on the contrast in seed size between paired seeds. The contrasts of tannin and nutrient content between paired seeds showed relatively few neighbor effects. Our results suggest that the contrast in seed traits between the target seed and its neighboring seeds should be considered when studying rodent-seed interactions. Furthermore, we expect that similar complex neighbor effects may also exist in other plant-animal interactions, such as pollination and herbivory.

Attitude Detection Classification based on Feature Selection and Integrated Machine Learning Model

This paper classifies 19 human actions based on a dataset of 1.2 million human actions acquired from sensors. The Xgboost classification model is established, and the MIV algorithm is used as an index to evaluate the importance of each variable to the dependent variable. The MIV values of all features are sorted according to the absolute value of each variable, and finally the top 10 groups of features are selected as the features of the reduced data set, and fine-tuned by grid search. Select the simplified data set, get the maximum roc_auc through continuous testing, and get the optimal model. The model has a recall of 1 and a precision, F1-score, and AUC of 0.99. Then, in order to make the model have a good generalization ability under the limited data set, a feasible method is designed to evaluate the generalization ability of the model. using the SMOTE-Tomek integrated sampling method to calculate k-nearest neighbor samples for each minority class sample, select the class samples whose neighbor similarity coefficients meet the requirements. Randomly generate new samples according to the adjacency relationship between the linearly interpolated sample and its neighbor samples. According to the original data training set, generate similar data samples and put them into the model trained with the original data for prediction evaluation. Finally, the classification accuracy rate of the generated samples is obtained, the precision rate is 0.98, and the recall rate is 0.99. The F1 value is 0.98, and the Roc_AUC value is 0.98, which proves that the evaluation model has good generalization ability.

ProsmORF-pred: a machine learning-based method for the identification of small ORFs in prokaryotic genomes.

Small open reading frames (smORFs) encoding proteins less than 100amino acids (aa) are known to be important regulators of key cellular processes. However, their computational identification remains a challenge. Based on a comprehensive analysis of known prokaryotic small ORFs, we have developed the ProsmORF-pred resource which uses a machine learning (ML)-based method for prediction of smORFs in the prokaryotic genome sequences. ProsmORF-pred consists of two ML models, one for initiation site recognition in nucleic acid sequences upstream of putative start codons and the other uses translated amino acid sequences to decipher functional protein like sequences. The nucleotide sequence-based initiation site recognition model has been trained using longer ORFs (>100aa) in the same genome while the ML model for identification of protein like sequences has been trained using annotated smORFs from Escherichia coli. Comprehensive benchmarking of ProsmORF-pred reveals that its performance is comparable to other state-of-the-art approaches on the annotated smORF set derived from 32 prokaryotic genomes. Its performance is distinctly superior to other tools like PRODIGAL and RANSEPS for prediction of newly identified smORFs which have a length range of 10-30aa, where prediction of smORFs has been a major challenge. Apart from identification of smORFs in genomic sequences, ProsmORF-pred can also aid in functional annotation of the predicted smORFs based on sequence similarity and genomic neighbourhood similarity searches in ProsmORFDB, a well-curated database of known smORFs. ProsmORF-pred along with its backend database ProsmORFDB is available as a user-friendly web server (http://www.nii.ac.in/prosmorfpred.html).

Similarity-navigated graph neural networks for node classification

Graph Neural Networks are effective in learning representations of graph-structured data. Some recent works are devoted to addressing heterophily, which exists ubiquitously in real-world networks, breaking the homophily assumption that nodes belonging to the same class are more likely to be connected and restricting the generalization of traditional methods in tasks such as node classification. However, these heterophily-oriented methods still lose efficacy in some typical heterophilic datasets. Moreover, issues on leveraging the knowledge from both node features and graph structure and investigating inherent properties of the datasets still need further consideration. In this work, we first provide insights based on similarity metrics to interpret the long-existing confusion that simple models sometimes perform better than models dedicated to heterophilic networks. Then, sticking to these insights and the classification principle of narrowing the intra-class distance and enlarging the inter-class distance of the sample's embeddings, we propose a Similarity-Navigated Graph Neural Network (SNGNN) which uses Node Similarity matrix coupled with mean aggregation operation instead of the normalized adjacency matrix in the neighborhood aggregation process. Moreover, based on SNGNN, a novel explicitly aggregating mechanism for selecting similar neighbors, named SNGNN+, is devised to preserve distinguishable features and handle the heterophilic problem. Additionally, a variant, SNGNN++, is further designed to adaptively integrate the knowledge from both node features and graph structure for improvement. Extensive experiments are conducted and demonstrate that our proposed framework outperforms the state-of-the-art methods for both small-scale and large-scale graphs regardless of their heterophilic extent. Our implementation is available online.

Diffusion Alignment Coefficient (DAC): A Novel Similarity Metric for Protein-Protein Interaction Network.

Interaction networks can be used to predict the functions of unknown proteins using known interactions and proteins with known functions. Many graph theory or diffusion-based methods have been proposed, using the assumption that the topological properties of a protein in a network are related to its biological function. Here we seek to improve function prediction by finding more similar neighbors with a new diffusion-based alignment technique to overcome the topological information loss of the node. In this study, we introduce the Diffusion Alignment Coefficient (DAC) algorithm, which combines diffusion, longest common subsequence, and longest common substring techniques to measure the similarity of two nodes in protein interaction networks. As a proof of concept, our experiments, conducted on a real PPI networks S.cerevisiae and Homo Sapiens, demonstrated that our method obtained better results than competitors for MIPS and MSigDB Collections hallmark gene set functional categories. This is the first study to develop a measure of node function similarity using alignment to consider the positions of nodes in protein-protein interaction networks. According to the experimental results, the use of spatial information belonging to the nodes in the network has a positive effect on the detection of more functionally similar neighboring nodes.

CAML: Contextual augmented meta-learning for cold-start recommendation

The performance of recommendation engines is challenged by the cold-start issues. Classical recommendation techniques have limited capability to address this problem because the underlying machine learning models are data-thirsty. Conversely, the recent success of meta-learning attracted a lot of attention due to few-shot learning capabilities which can be employed to accommodate new tasks even when there are few data samples for training. Most of the meta-learning recommendation models adopt model-agnostic meta-learning to initialize parameters that may lead to stuck into local optima instead of global optima for some users. To leverage the learning process, we propose a Contextually Augmented Meta-Learning recommender system (CAML). The proposed method augments the contextual features into a meta-learning model which considerably improves the tasks adaptation capability. We constructed a Data Augmentation Unit (DAU) that used a hybrid similarity to augment data samples of similar neighbors. The augmented samples are then forwarded to a Meta-Learner (MetaL) to learn user preferences and generate relevant recommendations. Additionally, we highlighted that the proposed method is generalized and can be adapted for various meta-learning-based recommendation models. We validate our proposal on three benchmark datasets for both ranking and rating prediction perspectives. The experimental outcomes prove the significance of our proposal over the state-of-the-art recommendation methods.

Density peaks clustering algorithm based on fuzzy and weighted shared neighbor for uneven density datasets

Uneven density data refers to data with a certain difference in sample density between clusters. The local density of density peaks clustering algorithm (DPC) does not consider the effect of sample density difference between clusters of uneven density data, which may lead to wrong selection of cluster centers; the algorithm allocation strategy makes it easy to incorrectly allocate samples originally belonging to sparse clusters to dense clusters, which reduces clustering efficiency. In this study, we proposed the density peaks clustering algorithm based on fuzzy and weighted shared neighbor for uneven density datasets (DPC-FWSN). First, a nearest neighbor fuzzy kernel function is obtained by combining K-nearest neighbor and fuzzy neighborhood. Then, local density is redefined by the nearest neighbor fuzzy kernel function. The local density can better characterize the distribution characteristics of the sample by balancing the contribution of sample density in dense and sparse areas, in order to avoid the situation that the sparse cluster does not have a cluster center. Finally, the allocation strategy for weighted shared neighbor similarity is proposed to optimize the sample allocation at the boundary of the sparse cluster. Experiments are performed on IDPC-FA, FKNN-DPC, FNDPC, DPCSA and DPC for uneven density datasets, complex morphologies datasets and real datasets. The clustering results demonstrate that DPC-FWSN effectively handles datasets with uneven density distribution.

Multi-channel nuclear norm minus Frobenius norm minimization for color image denoising

Color image denoising is frequently encountered in various image processing and computer vision tasks. One traditional strategy is to convert the RGB image to a less correlated color space and denoise each channel of the new space separately. However, such a strategy can not fully exploit the correlated information between channels and is inadequate to obtain satisfactory results. To address this issue, this paper proposes a new multi-channel optimization model for color image denoising under the nuclear norm minus Frobenius norm minimization framework. Specifically, based on the block-matching, the color image is decomposed into overlapping RGB patches. For each patch, we stack its similar neighbors to form the corresponding patch matrix. The proposed model is performed on the patch matrix to recover its noise-free version. During the recovery process, a) a weight matrix is introduced to fully utilize the noise difference between channels; b) the singular values are shrunk adaptively without additionally assigning weights. With them, the proposed model can achieve promising results while keeping simplicity. To solve the proposed model, an accurate and effective algorithm is built based on the alternating direction method of multipliers (ADMM) framework. The solution of each updating step can be analytically expressed in closed-from. Rigorous theoretical analysis proves that the solution sequences generated by the proposed algorithm converge to their respective stationary points. Experimental results on both synthetic and real noise data sets demonstrate the proposed model outperforms state-of-the-art models. MATLAB code is available at https://www.github.com/wangzhi-swu/MCNNFNM.

Ultra-fast meta-parameter optimization for time series similarity measures with application to nearest neighbour classification

Nearest neighbour similarity measures are widely used in many time series data analysis applications. They compute a measure of similarity between two time series. Most applications require tuning of these measures’ meta-parameters in order to achieve good performance. However, most measures have at least O(L^2) complexity, making them computationally expensive and the process of learning their meta-parameters burdensome, requiring days even for datasets containing only a few thousand series. In this paper, we propose UltraFastMPSearch, a family of algorithms to learn the meta-parameters for different types of time series distance measures. These algorithms are significantly faster than the prior state of the art. Our algorithms build upon the state of the art, exploiting the properties of a new efficient exact algorithm which supports early abandoning and pruning for most time series distance measures. We show on 128 datasets from the UCR archive that our new family of algorithms are up to an order of magnitude faster than the previous state of the art.