Adaptive Neighbor Research Articles

Training sample refining method using an adaptive neighbor to improve the classification performance of very high-spatial resolution remote sensing images

A training sample refining method is proposed to improve the classification performance of very high-spatial resolution (VHR) remote sensing images. The proposed approach involves three major steps. First, for a given image, an initial sample set with a limited number for each class is prepared manually. Second, neighboring pixels around each available labeled pixel are gradually distinguished by an adaptive extension algorithm. When an iterative extension around the available pixel is terminated, the neighboring pixels that are within the extended region are taken into account as candidate training samples. The candidate training sample is then used to refine the signature of each initial sample. Third, when the whole available labeled pixels are scanned and processed pixel-by-pixel in the above manner, the revised training sample set is trained specially for a supervised classifier for classification. Three VHR remote sensing images with limited initial samples are used for evaluating different classifiers and advanced methods based on spatial–spectral features to investigate the feasibility and performance of the proposed approach. Higher classification performance and accuracies are obtained by our proposed approach with respect to the classification maps based on the initial training sample set and an existing method that improves the initial training set by a regular window.

Hybrid QPSO-NNIA2 Algorithm for Multi-Objective Optimization Problem

To improve the convergence and distribution of a multi-objective optimization algorithm, a hybrid multi-objective optimization algorithm, based on the quantum particle swarm optimization (QPSO) algorithm and adaptive ranks clone and neighbor list-based immune algorithm (NNIA2), is proposed. The contribution of this work is threefold. First, the vicinity distance was used instead of the crowding distance to update the archived optimal solutions in the QPSO algorithm. The archived optimal solutions are updated and maintained by using the dynamic vicinity distance based m-nearest neighbor list in the QPSO algorithm. Secondly, an adaptive dynamic threshold of unfitness function for constraint handling is introduced in the process. It is related to the evolution algebra and the feasible solution. Thirdly, a new metric called the distribution metric is proposed to depict the diversity and distribution of the Pareto optimal. In order to verify the validity and feasibility of the QPSO-NNIA2 algorithm, we compare it with the QPSO, NNIA2, NSGA-II, MOEA/D, and SPEA2 algorithms in solving unconstrained and constrained multi-objective problems. The simulation results show that the QPSO-NNIA2 algorithm achieves superior convergence and superior performance by three metrics compared to other algorithms.

A Deep Learning Approach on Industrial Pyrolysis Reactor Monitoring

The pyrolysis process monitoring is always challenging due to the high operating temperature inside a fired furnace. To obtain better understanding of the pyrolysis reactors, we proposed a monitoring framework that builds upon thermal photography to provide a detailed view inside the fired furnace. Based on the infrared photos, the convolutional neural network is introduced into the monitoring framework to automatically recognize tube regions from the photos. In this work, a segmentation network is proposed based on the U-Net and ResNet-50 frameworks, by which the precise temperature and shape information on tube regions can be extracted from the raw photos. After extracting the important monitoring measurements, a control limit is drawn by the adaptive k-nearest neighbor method to detect abnormal conditions. The testing result indicates that the proposed monitoring framework provides in-depth information of the reactor and detailed fault diagnosis to process operators.

Diagnosis of gastric endoscopic biopsies using attenuated total reflectance (ATR) Fourier transform infrared (FT-IR) spectroscopy with entropy weight local-hyperplane k-nearest neighbor based on frequency domain information (EWHFI)

Fourier transform infrared spectroscopy (FT-IR) is an effective diagnostic tool for the characterization of biological molecules that may provide uninvasive, rapid detection of gastric cancer in vivo in the future. FT-IR spectra were obtained from 103 endoscopic biopsy gastric tissue samples that included 20 healthy, 35 chronic atrophic gastritis, 29 chronic superficial gastritis, and 19 gastric cancerous samples. A new approach using entropy weight local-hyperplane k-nearest neighbor based on frequency domain information (EWHFI) is proposed that improves and extends the adaptive weight local-hyperplane k-nearest neighbor (AWHK) approach. EWHFI combines the AWHK algorithm with the fine Fourier transform information in the frequency domain to effectively distinguish similar pathological states. The experimental results show that EWHFI enhances the accuracy, sensitivity, and specificity for the diagnosis of early gastric cancer. The average results of the random classification showed that the EWHFI classifiers for differentiating gastric cancerous, chronic atrophic gastritis, chronic superficial gastritis, and healthy tissues were 95.1%, 86.4%, 88.3%, and 99%, respectively.

LiDAR-Based Real-Time Detection and Modeling of Power Lines for Unmanned Aerial Vehicles.

The effective monitoring and maintenance of power lines are becoming increasingly important due to a global growing dependence on electricity. The costs and risks associated with the traditional foot patrol and helicopter-based inspections can be reduced by using UAVs with the appropriate sensors. However, this implies developing algorithms to make the power line inspection process reliable and autonomous. In order to overcome the limitations of visual methods in the presence of poor light and noisy backgrounds, we propose to address the problem of power line detection and modeling based on LiDAR. The PLDM, Power Line LiDAR-based Detection and Modeling, is a novel approach to detect power lines. Its basis is a scan-by-scan adaptive neighbor minimalist comparison for all the points in a point cloud. The power line final model is obtained by matching and grouping several line segments, using their collinearity properties. Horizontally, the power lines are modeled as a straight line, and vertically as a catenary curve. Using a real dataset, the algorithm showed promising results both in terms of outputs and processing time, adding real-time object-based perception capabilities for other layers of processing.

Unsupervised Learning of Human Pose Distance Metric via Sparsity Locality Preserving Projections

Human poses admit complicated articulations and multigranular similarity. Previous works on learning human pose metric utilize sparse models, which concentrate large weights on highly similar poses and fail to depict an overall structure of poses with multigranular similarity. Moreover, previous works require a large number of similar/dissimilar annotated pairwise poses, which is an tedious task and remains inaccurate due to different subjective judgments of experts. Motivated by graph-based neighbor assignment techniques, we propose an unsupervised model called sparsity locality preserving projection with adaptive neighbors (SLPPAN), for learning human pose distance metric. By using a property of the graph Laplacian, SLPPAN introduces a fixed-rank constraint to enforce an adaptive graph structure of poses and learns the neighbor assignment, the similarity measurement, and pose metric simultaneously. Experiments on pose retrieval of the CMU Mocap database demonstrate that SLPPAN outperforms traditional pose metric learning methods by capturing viewpoint variations of human poses. Experiments on keyframe extraction of the MSRAction3D database demonstrate that SLPPAN outperforms current methods by precisely detecting important frames of action sequences.

Scalable and Flexible Unsupervised Feature Selection.

Recently, graph-based unsupervised feature selection algorithms (GUFS) have been shown to efficiently handle prevalent high-dimensional unlabeled data. One common drawback associated with existing graph-based approaches is that they tend to be time-consuming and in need of large storage, especially when faced with the increasing size of data. Research has started using anchors to accelerate graph-based learning model for feature selection, while the hard linear constraint between the data matrix and the lower-dimensional representation is usually overstrict in many applications. In this letter, we propose a flexible linearization model with anchor graph and -norm regularization, which can deal with large-scale data sets and improve the performance of the existing anchor-based method. In addition, the anchor-based graph Laplacian is constructed to characterize the manifold embedding structure by means of a parameter-free adaptive neighbor assignment strategy. An efficient iterative algorithm is developed to address the optimization problem, and we also prove the convergence of the algorithm. Experiments on several public data sets demonstrate the effectiveness and efficiency of the method we propose.

A randomized adaptive neighbor discovery for wireless networks with multi-packet reception capability

Neighbor discovery is a first step in the initialization of wireless networks in large-scale ad hoc networks. In this paper, we propose a randomized neighbor discovery scheme for wireless networks with a multi-packet reception (MPR) capability. We let the nodes to have different advertisement probabilities. The node gradually adjusts its probability according to its operation phases: greedy or slow-start. In the greedy phase, the node advertises aggressively while it does moderately in the slow-start phase. Initial phase and advertisement probability are determined randomly. Then, the nodes change the probability adaptively according to advertisements the reception state from the other nodes. In order to decide the reception state precisely, the exact number of nodes in the network is necessary. To make our proposed scheme work in case of no prior knowledge of the population, we propose a population estimation method based on a maximum likelihood estimation. We evaluate our proposed scheme through numerical analysis and simulation. Through the numerical analysis, we show that the discovery completion time is lower bounded in Θ(Nk) and upper bounded in Θ(NlnNk) when there exists N nodes with MPR-k capability. The bounds are the same as those of previous studies that propose static optimal advertisement probability. Through the simulation, we evaluate that our adaptive scheme outperforms in terms of discovery completion time, advertisement efficiency, and wasted time slot ratio than a scheme with static advertisement probability when the population of the network is unknown.

Spectral Embedded Adaptive Neighbors Clustering.

Spectral clustering has been widely used in various aspects, especially the machine learning fields. Clustering with similarity matrix and low-dimensional representation of data is the main reason of its promising performance shown in spectral clustering. However, such similarity matrix and low-dimensional representation directly derived from input data may not always hold when the data are high dimensional and has complex distribution. First, the similarity matrix simply based on the distance measurement might not be suitable for all kinds of data. Second, the low-dimensional representation might not be able to reflect the manifold structure of the original data. In this brief, we propose a novel linear space embedded clustering method, which uses adaptive neighbors to address the above-mentioned problems. Linearity regularization is used to make the data representation a linear embedded spectral. We also use adaptive neighbors to optimize the similarity matrix and clustering results simultaneously. Extensive experimental results show promising performance compared with the other state-of-the-art algorithms.

An Improved Fuzzy K-Nearest Neighbor Algorithm for Imbalanced Data using Adaptive Approach

ABSTRACTFuzzy classification is a widely explored research solution of objects in data sciences and engineering. With the span of time, it got new heights with significant improvements according to the needs. Still there are some issues to be discussed and solved in a fuzzy manner; fuzzy classification of imbalanced data is one of them. Consequently, the importance of fuzzy nearest neighbor came into the scenario and deployed in many applications. Various improved crisp nearest neighbor approaches are performing well on imbalanced data-sets, but not much work has done on the fuzzy nearest neighbor for imbalanced data. In this paper, we propose to find out correct memberships of test instances from imbalanced data by merging an adaptive K-nearest neighbor approach to deal with the imbalanced issue and then join it with fuzzy K-nearest neighbor.

Fast unsupervised feature selection with anchor graph and ℓ2,1-norm regularization

Graph-based unsupervised feature selection has been proven to be effective in dealing with unlabeled and high-dimensional data. However, most existing methods face a number of challenges primarily due to their high computational complexity. In light of the ever-increasing size of data, these approaches tend to be inefficient in dealing with large-scale data sets. We propose a novel approach, called Fast Unsupervised Feature Selection (FUFS), to efficiently tackle this problem. Firstly, an anchor graph is constructed by means of a parameter-free adaptive neighbor assignment strategy. Meanwhile, an approximate nearest neighbor search technique is introduced to speed up the anchor graph construction. The l2,1-norm regularization is then performed to select more valuable features. Experiments on several large-scale data sets demonstrate the effectiveness and efficiency of the proposed method.

Anole: An Adaptive Neighbor Discovery Under Urban Environments

The fundamental operation in mobile wireless communication is to establish links between neighbor devices. The neighor discovery problem is non-trivial, especially under an urban environment. Various background scenarios, e.g., inside vehicles, at open-air squares, and in the supermarket, lead to complex discovery requirements. Discovery among fast moving devices requires an immediate exchange of emergency messages (minimum latency), while low-speed devices in crowded environments pay more attention to energy efficiency. Typical neighbor discovery protocols give solutions in a relatively stable scenario, which are not suited for the different environments in urban life. In this paper, we first proposed a non-integer framework to include all existing protocols. Then, a decentralized adaptive neighbor discovery protocol, named Anole , was designed under the framework. The protocol leveraged the genetic and similarity algorithms to be aware of and adapt to various scenarios with an appropriate discovery strategy. In the evaluation, we builtbuild different urban scenarios with real taxi and transportation card datasets in Shanghai. Meanwhile, an NS-3 simulator is applied to model device mobility and wireless communication. From the results, our protocol discovered 19 % more links with similar energy consumption than typical protocols.

Ranking with adaptive neighbors

Retrieving the most similar objects in a large-scale database for a given query is a fundamental building block in many application domains, ranging from web searches, visual, cross media, and document retrievals. State-of-the-art approaches have mainly focused on capturing the underlying geometry of the data manifolds. Graph-based approaches, in particular, define various diffusion processes on weighted data graphs. Despite success, these approaches rely on fixed-weight graphs, making ranking sensitive to the input affinity matrix. In this study, we propose a new ranking algorithm that simultaneously learns the data affinity matrix and the ranking scores. The proposed optimization formulation assigns adaptive neighbors to each point in the data based on the local connectivity, and the smoothness constraint assigns similar ranking scores to similar data points. We develop a novel and efficient algorithm to solve the optimization problem. Evaluations using synthetic and real datasets suggest that the proposed algorithm can outperform the existing methods.

Hybrid Amplitude Control Antenna Array System for Energy Efficient W-CDMA Network with Adaptive Coverage and Capacity

This paper introduces a novel technique and mathematical formulation of two dimensional elastic footprint variations and adaptive sector configuration variations by using a hybrid adaptive amplitude control antenna array system. The antenna array system can control the elevation beam-widths from 6.84° to 13° and the azimuth beam-widths from 33° to 93° for obtaining the adaptive capacity and coverage. The work illustrates the application of the antenna array system in the energy efficient W-CDMA network. The antenna array half-power beam width variation is achieved by using the antenna array amplitude control and co-located multiple antenna array switching. The Node-Bs can adaptively switch to three, four, five, six and non uniform sector configurations with the antenna array beam width control and adaptive neighbor relations. Sector configurations will vary according to the pre-defined specific traffic loading percentages out of Node-B’s total channel capacity to reduce the energy consumption of the entire network. The paper introduces an adaptive footprint overlaps shifting technique, where the sector with less traffic load will expand its footprint towards the sector with higher traffic load and the sector with higher traffic load will shrink its footprint by maintaining a fixed footprint overlaps to improve the coverage and the capacity of the network. The footprints overlap shifting is applicable only for the active sector antennas facing each other. The results confirm that the proposed antenna array system outperforms the automatic tilt optimization and can control coverage, capacity and improve the energy efficiency of the W-CDMA network.

A social recommendation method based on an adaptive neighbor selection mechanism

Recommender systems are techniques to make personalized recommendations of items to users. In e-commerce sites and online sharing communities, providing high quality recommendations is an important issue which can help the users to make effective decisions to select a set of items. Collaborative filtering is an important type of the recommender systems that produces user specific recommendations of the items based on the patterns of ratings or usage (e.g. purchases). However, the quality of predicted ratings and neighbor selection for the users are important problems in the recommender systems. Selecting suitable neighbors set for the users leads to improve the accuracy of ratings prediction in recommendation process. In this paper, a novel social recommendation method is proposed which is based on an adaptive neighbor selection mechanism. In the proposed method first of all, initial neighbors set of the users is calculated using clustering algorithm. In this step, the combination of historical ratings and social information between the users are used to form initial neighbors set for the users. Then, these neighbor sets are used to predict initial ratings of the unseen items. Moreover, the quality of the initial predicted ratings is evaluated using a reliability measure which is based on the historical ratings and social information between the users. Then, a confidence model is proposed to remove useless users from the initial neighbors of the users and form a new adapted neighbors set for the users. Finally, new ratings of the unseen items are predicted using the new adapted neighbors set of the users and the top_N interested items are recommended to the active user. Experimental results on three real-world datasets show that the proposed method significantly outperforms several state-of-the-art recommendation methods.

Concept Factorization With Adaptive Neighbors for Document Clustering

In this paper, a novel concept factorization (CF) method, called CF with adaptive neighbors (CFANs), is proposed. The idea of CFAN is to integrate an ANs regularization constraint into the CF decomposition. The goal of CFAN is to extract the representation space that maintains geometrical neighborhood structure of the data. Similar to the existing graph-regularized CF, CFAN builds a neighbor graph weights matrix. The key difference is that the CFAN performs dimensionality reduction and finds the neighbor graph weights matrix simultaneously. An efficient algorithm is also derived to solve the proposed problem. We apply the proposed method to the problem of document clustering on the 20 Newsgroups, Reuters-21578, and TDT2 document data sets. Our experiments demonstrate the effectiveness of the method.

Increasing Face Recognition Rates Using Novel Classification Algorithms

This paper describes and discusses a set of algorithms which can improve ace recognition rates. These algorithms include adaptive K-Nearest Neighbour, daptive weighted average, reverse weighted average and exponential weighted average. ssentially, the algorithms are extensions to the basic classification algorithm sed in most face recognition research. Whereas the basic classification algorithm elects the subject with the shortest associated distance, the algorithms presented in his paper manipulate and extract information from the set of distances between a est image and the training image set in order to obtain more accurate classifications. he base system to which the algorithms are applied uses the eigenfaces technique or recognition with an adapted Viola and Jones algorithm for face extraction. Most f the algorithms proposed show a consistent improvement over the baseline test.

Neighbor Discovery in Wireless Networks with Multipacket Reception

Neighbor discovery is one of the first steps in configuring and managing a wireless network. Most existing studies on neighbor discovery assume a single-packet reception model where only a single packet can be received successfully at a receiver. In this paper, motivated by the increasing prevalence of multipacket reception (MPR) technologies such as CDMA and MIMO, we study neighbor discovery in MPR networks that allow packets from multiple simultaneous transmitters to be received successfully at a receiver. Starting with a clique of n nodes, we first analyze a simple Aloha-like algorithm and show that it takes Θ((n ln n)/k) time to discover all neighbors with high probability when allowing up to k simultaneous transmissions. We then design two adaptive neighbor discovery algorithms that dynamically adjust the transmission probability for each node. We show that the adaptive algorithms yield a Θ(ln n) improvement over the Aloha-like scheme for a clique with n nodes and are thus order-optimal. Finally, we analyze our algorithms in a general multi-hop network setting. We show an upper bound of O((Δ ln n)/k) for the Aloha-like algorithm when the maximum node degree is Δ, which is at most a factor ln n worse than the optimal. In addition, when Δ is large, we show that the adaptive algorithms are orderoptimal, i.e., have a running time of O(Δ/k) which matches the lower bound for the problem.

An Adaptive Routing Protocol Based on QoS and Vehicular Density in Urban VANETs

Multihop data delivery between vehicles is an important technique to support the implementation of vehicular ad hoc networks (VANETs). However, many inherent characteristics of VANETs (e.g., dynamic network topology) bring great challenges to the data delivery. In particular, dynamic topology and intermittent connectivity make it difficult to design an efficient and stable geographic routing protocol for different applications of VANETs. To solve this problem, the paper proposes an adaptive routing protocol based on QoS and vehicular density (ARP-QD) in urban VANETs environments. The basic idea is to find the best path for end-to-end data delivery, which can satisfy diverse QoS requirements by considering hop count and link duration simultaneously. To reduce the network overhead furthermore, ARP-QD adopts an adaptive neighbor discovery algorithm to obtain neighbors’ information based on local vehicular density. In addition, a recovery strategy with carry-and-forward is utilized when the routing path is disrupted. Numerical simulations show that the proposed ARP-QD has higher delivery ratio than two prominent routing protocols in VANETs, without giving large compromise on delivery delay. The adaptivity of ARP-QD is also analyzed.

Using the global optimisation methods to minimise the machining path length of the free-form surfaces in three-axis milling

During the machining of free-form surfaces using three-axis numerically controlled machine (NC), several parameters are chosen arbitrary and one of the most important is the feed motion direction. The main objective of this study is to minimise the machining time of complex surfaces while respecting a scallop height criteria. The analytical expression of the machining time is not known and by hypothesis, it is assumed to be proportional to the path length crossed by the cutting tool. This path length depends on the feed direction. To have an optimal feed direction at any point, the surface is divided into zones with low variation of the steepest slope direction. The optimization problem was formulated aiming at minimizing the global path length. Furthermore, a penalty reflecting the time loss due to the movement of the tool from one zone to another one is taken into account. Several heuristics are used to resolve this problem: Clarke and Wrights, Greedy randomized adaptive search procedure, Tabu search and Nearest neighbour search. An example illustrates our work by applying the different heuristics on a test surface. After simulations, the results obtained present a significant saving of paths length of 24% compared to the machining in one zone.