Original Data Points Research Articles

Cross-modal hashing with semantic deep embedding

Cross-modal hashing has demonstrated advantages on fast retrieval tasks. It improves the quality of hash coding by exploiting semantic correlation across different modalities. In supervised cross-modal hashing, the learning of hash function replies on the quality of extracted features, for which deep learning models have been adopted to replace the traditional models based on handcraft features. All deep methods, however, have not sufficiently explored semantic correlation of modalities for the hashing process. In this paper, we introduce a novel end-to-end deep cross-modal hashing framework which integrates feature and hash-code learning into the same network. We take both between and within modalities data correlation into consideration, and propose a novel network structure and a loss function with dual semantic supervision for hash learning. This method ensures that the generated binary codes keep the semantic relationship of the original data points. Cross-modal retrieval experiments on commonly used benchmark datasets show that our method yields substantial performance improvement over several state-of-the-art hashing methods.

An efficient LSH indexing on discriminative short codes for high-dimensional nearest neighbors

In massive multimedia era, the dimension curse and the I/O performance bottleneck have become two major challenges for disk-based Approximate Nearest Neighbor (ANN) search. Hashing is a popular solution to overcome the dimension curse, one promising hashing technique is Locality Sensitive Hashing (LSH). However, most existing LSH indexings incur significant I/O cost during the search due to their low NN candidate hits in each I/O access. We recommend a novel method SC-LSH (SortingCodes-LSH) which combines LSH with another hashing technique (i.e., the discriminative short codes) to lift the hit of NN candidates so as to further boost the ANN search performance. Firstly, we intensify an LSH index and sort all the compound hashing keys according to a linear order to make similar NN candidates distributed locally. Then we generate product quantization (PQ) codes to use them as candidates instead of the original data points. These space-efficient short codes can enable us acquire significantly candidates via much less I/O operations. Moreover, based on theoretical and empirical studies among series of space-filling curves, we finally choose the Gray curve as the linear order to produce better local distribution of candidate data. All these above significantly increase the NN hits during each I/O, which greatly reduce the amount of necessary I/O access. Meanwhile, with the good similarity preserving ability, PQ codes are precise enough to discriminate NNs and thus guarantee the accuracy. Empirical study demonstrates that, comparing with four state-of-the-arts, SC-LSH achieves the best accuracy with significantly smaller I/O cost and space consumption. In fact, depending on the datasets, the I/O cost (resp., space consumption) of our scheme is only 5%-20% (resp., 1%-20%) of the other methods.

Multiview Latent Space Learning With Feature Redundancy Minimization.

Multiview learning has received extensive research interest and has demonstrated promising results in recent years. Despite the progress made, there are two significant challenges within multiview learning. First, some of the existing methods directly use original features to reconstruct data points without considering the issue of feature redundancy. Second, existing methods cannot fully exploit the complementary information across multiple views and meanwhile preserve the view-specific properties; therefore, the degraded learning performance will be generated. To address the above issues, we propose a novel multiview latent space learning framework with feature redundancy minimization. We aim to learn a latent space to mitigate the feature redundancy and use the learned representation to reconstruct every original data point. More specifically, we first project the original features from multiple views onto a latent space, and then learn a shared dictionary and view-specific dictionaries to, respectively, exploit the correlations across multiple views as well as preserve the view-specific properties. Furthermore, the Hilbert-Schmidt independence criterion is adopted as a diversity constraint to explore the complementarity of multiview representations, which further ensures the diversity from multiple views and preserves the local structure of the data in each view. Experimental results on six public datasets have demonstrated the effectiveness of our multiview learning approach against other state-of-the-art methods.

Effect of inclination on heat transfer coefficient during flow boiling in a mini-channel

Abstract An experimental study of heat transfer during R245fa flow boiling is carried out in a 1.6 mm inner diameter circular channel (198 original data points). The test section is composed of a sapphire tube coated with ITO, which enables a total transparency of the evaporator in the visible spectrum. The acquisition of the outer wall temperature field is ensured by means of an infrared camera and the heat transfer coefficient is calculated by means of a local thermal model. The effect of the inclination on the heat transfer is presented and discussed for various vapour qualities and mass velocities and a saturation temperature of 81 °C, corresponding to a Bond number of 4.1. For each experimental condition, the mean heat transfer coefficient is calculated for various inclinations from the vertical downward flow (−90°) to the vertical upward flow (+90°). The variation of the heat transfer coefficient are first presented and analyzed in the horizontal case. The heat transfer coefficient variations with the flow parameters are interpreted in terms of nucleate and convective boiling and these observations are confirmed by the visualisation of the nucleation phenomenon in the evaporator. The inclination has almost no effect on the heat transfer and this lack of effect is finally interpreted in terms of shift between the annular to intermittent flow pattern transition coupled with the nucleate to convective boiling transition.

Optimizing the Shape Parameters of Beta-Spline Using Particle Swarm Optimization

Beta-spline is an alternative curve for 2D font representation. It is preferred since it has G2 continuity and two shape parameters, that can be used to control the curve shape. These shape parameters can also be used to optimize the error between fitted curve and original data points. Commonly, most of the researcher use value of shape parameters of beta-spline as and or some of the researcher choose any random value of these two shape parameters that suitable to be used in beta-spline curve fitting. The values of shape parameters are very important since the values affect the total error of the fitted curves. Thus, in this paper, Particle Swarm Optimization (PSO) is employed to determine the optimum value of the two shape parameters that will optimize the approximation error of the fitted curve. The technique is applied on two fonts: ى and δ, and tested using various number of iterations and populations.

Automated detection of imaging features of disproportionately enlarged subarachnoid space hydrocephalus using machine learning methods.

ObjectiveCreate an automated classifier for imaging characteristics of disproportionately enlarged subarachnoid space hydrocephalus (DESH), a neuroimaging phenotype of idiopathic normal pressure hydrocephalus (iNPH).Methods1597 patients from the Mayo Clinic Study of Aging (MCSA) were reviewed for imaging characteristics of DESH. One core feature of DESH, the presence of tightened sulci in the high-convexities (THC), was used as a surrogate for the presence of DESH as the expert clinician-defined criterion on which the classifier was trained. Anatomical MRI scans were automatically segmented for cerebrospinal fluid (CSF) and overlaid with an atlas of 123 named sulcal regions. The volume of CSF in each sulcal region was summed and normalized to total intracranial volume. Area under the receiver operating characteristic curve (AUROC) values were computed for each region individually, and these values determined feature selection for the machine learning model. Due to class imbalance in the data (72 selected scans out of 1597 total scans) adaptive synthetic sampling (a technique which generates synthetic examples based on the original data points) was used to balance the data. A support vector machine model was then trained on the regions selected.ResultsUsing the automated classification model, we were able to classify scans for tightened sulci in the high convexities, as defined by the expert clinician, with an AUROC of about 0.99 (false negative ≈ 2%, false positive ≈ 5%). Ventricular volumes were among the classifier's most discriminative features but are not specific for DESH. The inclusion of regions outside the ventricles allowed specificity from atrophic neurodegenerative diseases that are also accompanied by ventricular enlargement.ConclusionAutomated detection of tight high convexity, a key imaging feature of DESH, is possible by using support vector machine models with selected sulcal CSF volumes as features.

Non-Contact Body Measurement for Qinchuan Cattle with LiDAR Sensor.

The body dimension measurement of large animals plays a significant role in quality improvement and genetic breeding, and the non-contact measurements by computer vision-based remote sensing could represent great progress in the case of dangerous stress responses and time-costing manual measurements. This paper presents a novel approach for three-dimensional digital modeling of live adult Qinchuan cattle for body size measurement. On the basis of capturing the original point data series of live cattle by a Light Detection and Ranging (LiDAR) sensor, the conditional, statistical outliers and voxel grid filtering methods are fused to cancel the background and outliers. After the segmentation of K-means clustering extraction and the RANdom SAmple Consensus (RANSAC) algorithm, the Fast Point Feature Histogram (FPFH) is put forward to get the cattle data automatically. The cattle surface is reconstructed to get the 3D cattle model using fast Iterative Closest Point (ICP) matching with Bi-directional Random K-D Trees and a Greedy Projection Triangulation (GPT) reconstruction method by which the feature points of cattle silhouettes could be clicked and calculated. Finally, the five body parameters (withers height, chest depth, back height, body length, and waist height) are measured in the field and verified within an accuracy of 2 mm and an error close to 2%. The experimental results show that this approach could be considered as a new feasible method towards the non-contact body measurement for large physique livestock.

Unified Simultaneous Clustering and Feature Selection for Unlabeled and Labeled Data.

This paper proposes a novel feature selection method, namely, unified simultaneous clustering feature selection (USCFS). A regularized regression with a new type of target matrix is formulated to select the most discriminative features among the original features from labeled or unlabeled data. The regression with -norm regularization allows the projection matrix to represent an effective selection of discriminative features. For unsupervised feature selection, the target matrix discovers label-like information not from the original data points but rather from projected data points, which are of a reduced dimensionality. Without the aid of an affinity graph-based local structure learning method, USCFS allows the target matrix to capture latent cluster centers via orthogonal basis clustering and to simultaneously select discriminative features guided by latent cluster centers. When class labels are available, the target matrix is also able to find latent class labels by regarding the ground-truth class labels as an approximate guide. Hence, supervised feature selection is realized using these latent class labels, which may differ from the ground-truth class labels. Experimental results demonstrate the effectiveness of the proposed method. Specifically, the proposed method outperforms the state-of-the-art methods on diverse real-world data sets for both the supervised and the unsupervised feature selection.

Quality Control Procedure Based on Partitioning of NMR Time Series.

The quality of the magnetic resonance spectroscopy (MRS) depends on the stability of magnetic resonance (MR) system performance and optimal hardware functioning, which ensure adequate levels of signal-to-noise ratios (SNR) as well as good spectral resolution and minimal artifacts in the spectral data. MRS quality control (QC) protocols and methodologies are based on phantom measurements that are repeated regularly. In this work, a signal partitioning algorithm based on a dynamic programming (DP) method for QC assessment of the spectral data is described. The proposed algorithm allows detection of the change points—the abrupt variations in the time series data. The proposed QC method was tested using the simulated and real phantom data. Simulated data were randomly generated time series distorted by white noise. The real data were taken from the phantom quality control studies of the MRS scanner collected for four and a half years and analyzed by LCModel software. Along with the proposed algorithm, performance of various literature methods was evaluated for the predefined number of change points based on the error values calculated by subtracting the mean values calculated for the periods between the change-points from the original data points. The time series were checked using external software, a set of external methods and the proposed tool, and the obtained results were comparable. The application of dynamic programming in the analysis of the phantom MRS data is a novel approach to QC. The obtained results confirm that the presented change-point-detection tool can be used either for independent analysis of MRS time series (or any other) or as a part of quality control.

A closed-form unsupervised geometry-aware dimensionality reduction method in the Riemannian Manifold of SPD matrices.

Riemannian geometry has been found accurate and robust for classifying multidimensional data, for instance, in brain-computer interfaces based on electroencephalography. Given a number of data points on the manifold of symmetric positive-definite matrices, it is often of interest to embed these points in a manifold of smaller dimension. This is necessary for large dimensions in order to preserve accuracy and useful in general to speed up computations. Geometry-aware methods try to accomplish this task while respecting as much as possible the geometry of the original data points. We provide a closed-form solution for this problem in a fully unsupervised setting. Through the analysis of three brain-computer interface data bases we show that our method allows substantial dimensionality reduction without affecting the classification accuracy.

Functional analysis of high-content high-throughput imaging data

ABSTRACTHigh-content automated imaging platforms allow the multiplexing of several targets simultaneously to generate multi-parametric single-cell data sets over extended periods of time. Typically, standard simple measures such as mean value of all cells at every time point are calculated to summarize the temporal process, resulting in loss of time dynamics of the single cells. Multiple experiments are performed but observation time points are not necessarily identical, leading to difficulties when integrating summary measures from different experiments. We used functional data analysis to analyze continuous curve data, where the temporal process of a response variable for each single cell can be described using a smooth curve. This allows analyses to be performed on continuous functions, rather than on original discrete data points. Functional regression models were applied to determine common temporal characteristics of a set of single cell curves and random effects were employed in the models to explain variation between experiments. The aim of the multiplexing approach is to simultaneously analyze the effect of a large number of compounds in comparison to control to discriminate between their mode of action. Functional principal component analysis based on T-statistic curves for pairwise comparison to control was used to study time-dependent compound effects.

Accelerating flow propagator measurements for the investigation of reactive transport in porous media

NMR propagator measurements are widely used for identifying the distribution of molecular displacements over a given observation time, characterising a flowing system. However, where high q-space resolution is required, the experiments are time consuming and therefore unsuited to the study of dynamic systems. Here, it is shown that with an appropriately sampled subset of the q-space points in a high-resolution flow propagator measurement, one can quickly and robustly reconstruct the fully sampled propagator through interpolation of the acquired raw data. It was found that exponentially sampling ∼4% of the original data-points allowed a reconstruction with the deviation from the fully sampled propagator below the noise level, in this case reducing the required experimental time from ∼2.8h to <7min. As a demonstration, this approach is applied to observe the temporal evolution of the reactive flow of acid through an Estaillades rock core plug. It is shown that ‘wormhole’ formation in the rock core plug provides a channel for liquid flow such that the remaining pore space is by-passed, thereby causing the flow velocity of the liquid in the remaining part of the plug to become stagnant. The propagator measurements are supported by both 1D profiles and 2D imaging data. Such insights are of importance in understanding well acidisation and CO2 sequestration processes.

Method for Determining Appropriate Clustering Criteria of Location-Sensing Data

Large quantities of location-sensing data are generated from location-based social network services. These data are provided as point properties with location coordinates acquired from a global positioning system or Wi-Fi signal. To show the point data on multi-scale map services, the data should be represented by clusters following a grid-based clustering method, in which an appropriate grid size should be determined. Currently, there are no criteria for determining the proper grid size, and the modifiable areal unit problem has been formulated for the purpose of addressing this issue. The method proposed in this paper is applies a hexagonal grid to geotagged Twitter point data, considering the grid size in terms of both quantity and quality to minimize the limitations associated with the modifiable areal unit problem. Quantitatively, we reduced the original Twitter point data by an appropriate amount using Töpfer’s radical law. Qualitatively, we maintained the original distribution characteristics using Moran’s I. Finally, we determined the appropriate sizes of clusters from zoom levels 9–13 by analyzing the distribution of data on the graphs. Based on the visualized clustering results, we confirm that the original distribution pattern is effectively maintained using the proposed method.

TiM: Fine-Grained Rate Adaptation in WLANs

Channel condition varies frequently in wireless networks. To achieve good performance, devices need rate adaptation. In rate adaptation, choosing proper modulation schemes based on channel conditions is vital to the transmission performance. However, due to the natural character of discrete modulation types and continuous varied link conditions, we cannot make a one-to-one mapping from modulation schemes to channel conditions. This matching gap causes either over-select or under-select modulation schemes which limits throughput performance. To fill-in the gap, we propose time-line modulation (TiM), a novel three-Dimensional modulation scheme by adding time dimension into current amplitude-phase domain schemes. With estimation of channel condition, TiM changes base-band data transmission time by artificially interpolating values between original data points without changing amplitude-phase domain modulation type. We implemented TiM on USRP2 and conducted comprehensive simulations. Results show that, compared with rate adaptation choosing from traditional modulation schemes, TiM can improve channel utilization up to 200 percent.

Adapting Hidden Markov Models for Online Learning

In modern computer systems, the intermittent behaviour of infrequent, additional loads affects performance. Often, representative traces of storage disks or remote servers can be scarce and obtaining real data is sometimes expensive. Therefore, stochastic models, through simulation and profiling, provide cheaper, effective solutions, where input model parameters are obtained. A typical example is the Markov-modulated Poisson process (MMPP), which can have its time index discretised to form a hidden Markov model (HMM). These models have been successful in capturing bursty behaviour and cyclic patterns of I/O operations and Internet traffic, using underlying properties of the discrete (or continuous) Markov chain. However, learning on such models can be cumbersome in terms of complexity through re-training on data sets. Thus, we provide an online learning HMM (OnlineHMM), which is composed of two existing variations of HMMs: first, a sliding HMM using a moving average technique to update its parameters “on-the-fly” and, secondly, a multi-input HMM capable of training on multiple discrete traces simultaneously. The OnlineHMM reduces data processing times significantly and thence synthetic workloads become computationally more cost effective. We measure the accuracy of reproducing representative traces through comparisons of moments and autocorrelation on original data points and HMM-generated synthetic traces. We present, analytically, the training steps saved through the OnlineHMM's adapted Baum-Welch algorithm and obtain, through simulation, mean waiting times of a queueing model. Finally, we conclude our work and offer model extensions for the future.

Diffusion in the Layered Li-Ni-Mn-Co Oxide Based Core-Shell/Gradient Materials during Sintering Simulated with Pellets in Contact

Core-shell structured positive electrode materials based on layered Li-Ni-Mn-Co oxide could be the next generation of positive electrode materials for high energy density lithium-ion batteries. This is because a high energy core material with poor stability against the electrolyte can be protected by a thin layer of a stable shell material. In our previous report1,2, Li and Mn-rich materials were used as the protecting shell, and Ni-rich materials were used as the core. It was shown that the Mn-rich shell can effectively protect the Ni-rich core from reactions with the electrolyte while the Ni-rich core renders a high and stable average voltage.2However, it was also found that diffusion of the cations between the core and shell phases occurs during sintering. Figure 1 shows energy dispersive X-ray spectroscopy (EDS) results of the core-shell precursor and lithiated samples. It is demonstrated that a Mn-rich shell was still maintained whereas the Co, which was only in the shell in the precursor, was approximately homogeneous throughout the particles after sintering at 900oC compared to the prepared hydroxide precursors. This diffusion phenomena in core-shell/gradient materials has not been previously examined in detail, to our knowledge, despite the large amount of work on the nickel rich NMC (LiNixMnyCozO2) (without excess lithium) gradient cathode material, which has been published and cited.3–5In this work, a series of experiments to measure the interdiffusivity of the cations were designed. Pellets of core material and shell material were placed in contact and heated to induce cation diffusion over long, measurable distances. Based on the analyzed composition profiles, interdiffusion constants can be extracted. Li1.09(Ni0.8Mn0.2)0.81O2, Li1.0 5(Co0.9Mn0.1)0.8 5O2, LiNi0.8Co0.2O2 and LiCoO2 powders were produced at 900oC in oxygen for 10 h, while Li2MnO3 was made in air. Powder a was first compressed to 15 MPa for 5 minutes, then powder b was added to the mould on top of the pellet a. The powders were compressed again to 50 MPa for 10 mins to ensure a good contact at the interface. The pellets were heated to 900oC for 10 h and then cooled slowly. The heated pellets were purposely fractured in to smaller pieces, which were subsequently embedded into Crystal BondTM, and eventually polished to a mirror surface finish for scanning electron microscopy (SEM) and EDS experiments. Figure 2 shows the back scattered SEM image of the cross-section of a pellet that had LiCoO2 powder on the left side and LiNi0.8Co0.2O2 on the right side. Small pores are observed at the interface. Figure 3a shows the Ni and Mn atomic concentration profile against position for the Li1.09(Ni0.8Mn0.2)0.81O2/Li2MnO3 pellet measured using EDS line scans. The symbols show the original data points. The concentration of Ni decreased from 80% to about 0% in about 2 µm while the concentration of Mn increased from 20% to about 100% over the same distance. Figure 3b shows the Ni and Co atomic concentration profiles versus position for the LiCoO2/LiNi0.8Co0.2O2 pellet measured using EDS. Figure 3b shows that the atomic concentration of Co decreased from 100% to about 20% in about 20 µm, which is about 10 times larger than that for the Ni and Mn shown in Figure 3a, indicating the interdiffusion between Ni3+ and Co3+ is much larger than that between Ni3+ and Mn4+. This result agrees well with the trends observed in the core-shell samples shown in Figure 1. Interdiffusion constants of Ni, Co and Mn in various couples as a function of temperature will be extracted and will be reported in the presentation.

A Method of DTM Construction Based on Quadrangular Irregular Networks and Related Error Analysis.

A new method of DTM construction based on quadrangular irregular networks (QINs) that considers all the original data points and has a topological matrix is presented. A numerical test and a real-world example are used to comparatively analyse the accuracy of QINs against classical interpolation methods and other DTM representation methods, including SPLINE, KRIGING and triangulated irregular networks (TINs). The numerical test finds that the QIN method is the second-most accurate of the four methods. In the real-world example, DTMs are constructed using QINs and the three classical interpolation methods. The results indicate that the QIN method is the most accurate method tested. The difference in accuracy rank seems to be caused by the locations of the data points sampled. Although the QIN method has drawbacks, it is an alternative method for DTM construction.

Bahadur representations for bootstrap quantiles

A bootstrap sample may contain more than one replica of original data points. To extend the classical Bahadur type representations for the sample quantiles in the independent identical distributed case to bootstrap sample quantiles therefore is not a trivial task. This manuscript fulfils the task and establishes the asymptotic theory of bootstrap sample quantiles.

Information visualization by dimensionality reduction: a review

Information visualization can be considered a process of transforming similarity relationships between data points to a geometric representation in order to see unseen information. High-dimensionality data sets are one of the main problems of information visualization. Dimensionality Reduction (DR) is therefore a useful strategy to project high-dimensional space onto low-dimensional space, which it can be visualized directly. The application of this technique has several benefits. First, DR can minimize the amount of storage needed by reducing the size of the data sets. Second, it helps to understand the data sets by discarding any irrelevant features, and to focus on the main important features. DR can enable the discovery of rich information, which assists the task of data analysis. Visualization of high-dimensional data sets is widely used in many fields, such as remote sensing imagery, biology, computer vision, and computer graphics. The visualization is a simple way to understand the high-dimensional space because the relationship between original data points is incomprehensible. A large number of DR methods which attempt to minimize the loss of original information. This paper discuss and analys some DR methods to support the idea of dimensionality reduction to get trustworthy visualization. Keywords: Dimensionality Reduction, Information visualization, Information retrieval.

Constraint neighborhood projections for semi-supervised clustering.

Semi-supervised clustering aims to incorporate the known prior knowledge into the clustering algorithm. Pairwise constraints and constraint projections are two popular techniques in semi-supervised clustering. However, both of them only consider the given constraints and do not consider the neighbors around the data points constrained by the constraints. This paper presents a new technique by utilizing the constrained pairwise data points and their neighbors, denoted as constraint neighborhood projections that requires fewer labeled data points (constraints) and can naturally deal with constraint conflicts. It includes two steps: 1) the constraint neighbors are chosen according to the pairwise constraints and a given radius so that the pairwise constraint relationships can be extended to their neighbors, and 2) the original data points are projected into a new low-dimensional space learned from the pairwise constraints and their neighbors. A CNP-Kmeans algorithm is developed based on the constraint neighborhood projections. Extensive experiments on University of California Irvine (UCI) datasets demonstrate the effectiveness of the proposed method. Our study also shows that constraint neighborhood projections (CNP) has some favorable features compared with the previous techniques.