Search

Background: Functional mapping, despite its proven efficiency, suffers from a “chicken or egg” sce- nario, in that, poor spatial features lead to inadequate spectral alignment and vice versa during training, often resulting in slow convergence, high computational costs, and learning failures, particularly when small datasets are used. Methods: A novel method is presented for dense-shape correspondence, whereby the spatial information transformed by neural networks is combined with the projections onto spectral maps to overcome the “chicken or egg” challenge by selectively sampling only points with high confidence in their alignment. These points then contribute to the alignment and spectral loss terms, boosting training, and accelerating convergence by a factor of five. To ensure full unsupervised learning, the Gromov–Hausdorff distance metric was used to select the points with the maximal alignment score displaying most confidence. Results: The effectiveness of the proposed approach was demonstrated on several benchmark datasets, whereby results were reported as superior to those of spectral and spatial-based methods. Conclusions: The proposed method provides a promising new approach to dense-shape correspondence, addressing the key challenges in the field and offering significant advantages over the current methods, including faster convergence, improved accuracy, and reduced computational costs.

Online active learning is a paradigm in machine learning that aims to select the most informative data points to label from a data stream. The problem of minimizing the cost associated with collecting labeled observations has gained a lot of attention in recent years, particularly in real-world applications where data is only available in an unlabeled form. Annotating each observation can be time-consuming and costly, making it difficult to obtain large amounts of labeled data. To overcome this issue, many active learning strategies have been proposed in the last decades, aiming to select the most informative observations for labeling in order to improve the performance of machine learning models. These approaches can be broadly divided into two categories: static pool-based and stream-based active learning. Pool-based active learning involves selecting a subset of observations from a closed pool of unlabeled data, and it has been the focus of many surveys and literature reviews. However, the growing availability of data streams has led to an increase in the number of approaches that focus on online active learning, which involves continuously selecting and labeling observations as they arrive in a stream. This work aims to provide an overview of the most recently proposed approaches for selecting the most informative observations from data streams in real time. We review the various techniques that have been proposed and discuss their strengths and limitations, as well as the challenges and opportunities that exist in this area of research. [ABSTRACT FROM AUTHOR]

In his target article on "Indoctrination in Introduction to Psychology," Bartels proposes that in introductory textbooks of psychology studies and their results are systematically presented in such a way that they tend to correspond to left-liberal political positions and that the state of psychological knowledge is reflected in a correspondingly distorted way. In our commentary, we clarify that the evidence Bartels presents for this claim is insufficient. At first, he takes a purely hypothesis-confirming approach based on selective sampling. Second, he draws an invalid causal inference from a supposed liberal majority in the psychological community to their representation of psychological content in textbooks. And third, he assigns introductory textbooks a function that we believe they do not have. Nonetheless, we welcome the discussion of how best to teach critical reflective thinking in psychology courses. [ABSTRACT FROM AUTHOR]

Subpixel-based image down-sampling has been widely used to improve the apparent resolution of down-sampled images on display. However, previous subpixel rendering methods often introduce distortions, such as aliasing and color-fringing. This study proposes a novel subpixel rendering method that uses selective sampling and optimal filtering. We first generalize the previous frequency domain analysis results indicating the relationships between various down-sampling patterns and the aliasing artifact. Based on this generalized analysis, a subpixel-based down-sampling pattern for each image is selectively determined by utilizing the edge distribution of the image. Moreover, we investigate the origin of the color-fringing artifact in the frequency domain. Optimal spatial filters that can effectively remove distortions caused by the selected down-sampling pattern are designed via frequency domain analyses of aliasing and color-fringing. The experimental results show that the proposed method is not only robust to the aliasing and color-fringing artifacts but also outperforms the existing ones in terms of information preservation.

There is increasing interest in methods to disentangle the relationship between genotype and (endo)phenotypes in human complex traits. We present a population-based method of increasing the power and cost-efficiency of studies by selecting random individuals with a particular genotype and then assessing the accompanying quantitative phenotypes. Using statistical derivations, power- and cost graphs we show that such a "forward genetics" approach can lead to a marked reduction in sample size and costs. This approach is particularly apt for implementing in epidemiological studies for which DNA is already available but the phenotyping costs are high.

Data often exists in subspaces embedded within a high-dimensional space. Subspace clustering seeks to group data according to the dimensions relevant to each subspace. This requires the estimation of subspaces as well as the clustering of data. Subspace clustering becomes increasingly challenging in high dimensional spaces due to the curse of dimensionality which affects reliable estimations of distances and density. Recently, another aspect of high-dimensional spaces has been observed, known as the hubness phenomenon, whereby few data points appear frequently as nearest neighbors of the rest of the data. The distribution of neighbor occurrences becomes skewed with increasing intrinsic dimensionality of the data, and few points with high neighbor occurrences emerge as hubs. Hubs exhibit useful geometric properties and have been leveraged for clustering data in the full-dimensional space. In this paper, we study hubs in the context of subspace clustering. We present new characterizations of hubs in relation to subspaces, and design graph-based meta-features to identify a subset of hubs which are well fit to serve as seeds for the discovery of local latent subspaces and clusters. We propose and evaluate a hubness-driven algorithm to find subspace clusters, and show that our approach is superior to the baselines, and is competitive against state-of-the-art subspace clustering methods. We also identify the data characteristics that make hubs suitable for subspace clustering. Such characterization gives valuable guidelines to data mining practitioners. [ABSTRACT FROM AUTHOR]

Objectives: Atypical hyperplasia of the endometrium is a significant risk factor for uterine endometrioid carcinoma (EC) and an indication for hysterectomy. Standard sampling of these specimens includes evaluation of the entire endometrium to identify possible EC. We evaluated a method of selective sampling in an effort to balance resource utilization with diagnostic accuracy in the detection of EC.Methods: Histologic diagnoses based on selective sampling (exclusion of every other block of endometrium) were compared with the original diagnosis based on complete sampling.Results: Double-blinded review of these cases using selective sampling detected EC in 92% of hysterectomies, including all high-grade/high-stage carcinomas. Selective sampling had an 82% agreement with the original diagnoses, with most discordant diagnoses attributable to interobserver variability. Adjusting for interobserver variability increased diagnostic agreement between selective and complete sampling to 96%.Conclusions: Selective sampling is a feasible method to save time and resources while maintaining diagnostic accuracy. [ABSTRACT FROM AUTHOR]

In this paper we describe selective sampling algorithms for decision trees and random forests and their contribution to the classification accuracy. In our selective sampling algorithms, the instance that yields the highest expected utility is chosen to be labeled by the expert. We show that it is possible to obtain the most valuable unlabeled instance to be labeled by the expert and added to the training dataset of the decision tree simply by depicting the influence of this new instance on the class probabilities of the leaves. All the unlabeled instances that fall into the same leaf will have the same class probabilities. As a result, we can compute the expected accuracy of the decision tree according to its leaves instead for each individual unlabeled instance. An extension for random forests is also presented. Moreover, we show that the selective sampling classifier has to belong to the same family as the classifier whose accuracy we wish to improve but need not be identical to it. For example, a random forest classifier can be used for the selective sampling process, and the results can be used to improve the classification accuracy of a decision tree. Likewise, a random forest classifier consisting of three trees can be used in the selective sampling algorithm to improve the classification accuracy of a random forest consisting of ten trees. Our experiments show that the proposed selective sampling algorithms achieve better accuracy than the standard random sampling, uncertainty sampling and the active belief decision tree learning approach (ABC4.5) for several real-world datasets. We also show that our selective sampling algorithms improve significantly the classification performance of several state-of-the-art classifiers such as the random rotation forest classifier for real-world large-scale datasets. [ABSTRACT FROM AUTHOR]

For obtaining a high-quality profile of measured sculptured surface, scanning devices have to produce massive point cloud data with great sampling rates. Bottlenecks are created owing to inefficiencies in storing, manipulating and transferring these data, and the parametric modeling from them is quite a time-consuming work. The purpose of this paper is to effectively simplify point cloud data from a measured sculptured surface during the on-line point cloud data selective sampling process. The key contribution is the generation of a novel reasoning mechanism which is based on a predictor–corrector scheme, and it is capable of eliminating data redundancy caused by spatial similarity of collected point clouds. In particular, this mechanism is embedded in our newly designed framework for on-line point cloud data selective sampling of sculptured surfaces. This framework consists of two stages: First, the initial point data flow is selective sampled using bi-Akima method; second, the data flow is refined based on our proposed reasoning mechanism. Moreover, our versatile framework is capable of obtaining high-quality resampling results with smaller data reduction ratio than other existing on-line point cloud data reduction/selective sampling methods. Experiment is conducted and the results demonstrate the superior performance of the proposed method. • A versatile framework for point cloud data reduction of sculptured surfaces is designed. • A novel reasoning mechanism is proposed and embedded in this framework. • It is able to eliminate data redundancy caused by spatial similarity of point clouds. • It gets high-quality resampling results with smaller data reduction ratio than other methods. • 3D scans from six simulated surfaces validate the proposed method. [ABSTRACT FROM AUTHOR]

A selective classifier (f, g) comprises a classification function f and a binary selection function g, which determines if the classifier abstains from prediction, or uses f to predict. The classifier is called pointwise-competitive if it classifies each point identically to the best classifier in hindsight (from the same class), whenever it does not abstain. The quality of such a classifier is quantified by its rejection mass, defined to be the probability mass of the points it rejects. A "fast" rejection rate is achieved if the rejection mass is bounded from above by ~O(1=m) where m is the number of labeled examples used to train the classi fier (and O hides logarithmic factors). Pointwise-competitive selective (PCS) classifiers are intimately related to disagreement-based active learning and it is known that in the realizable case, a fast rejection rate of a known PCS algorithm (called Consistent Selective Strategy) is equivalent to an exponential speedup of the well-known CAL active algorithm. We focus on the agnostic setting, for which there is a known algorithm called LESS that learns a PCS classifier and achieves a fast rejection rate (depending on Hanneke's disagreement coefficient) under strong assumptions. We present an improved PCS learning algorithm called ILESS for which we show a fast rate (depending on Hanneke's disagreement coefficient) without any assumptions. Our rejection bound smoothly interpolates the realizable and agnostic settings. The main result of this paper is an equivalence between the following three entities: (i) the existence of a fast rejection rate for any PCS learning algorithm (such as ILESS); (ii) a poly-logarithmic bound for Hanneke's disagreement coefficient; and (iii) an exponential speedup for a new disagreement-based active learner called Active-ILESS. [ABSTRACT FROM AUTHOR]

We study weakly-supervised anomaly detection where only video-level "anomalous"/"normal" labels are available in training, while anomaly events should be temporally localized in testing. For this task, a commonly used framework is multiple instance learning (MIL), where clip instances are sampled from individual videos to form video-level bags. This sampling process arguably is a bottleneck of MIL. If too many instances are sampled, we not only encounter high computational overheads but also have many noisy instances in the bag. On the other hand, when too few instances are used, e. g., through enlarged grids, much background noise may be included in the anomaly instances. To resolve this dilemma, we propose a simple yet effective method named Sub-Max. In partitioned image regions, it identifies instances that are most probable candidates for anomaly events by selecting cuboids that have high optical flow magnitudes. We show that our method effectively brings down the computational cost of the baseline MIL and at the same time significantly filters out the influence of noise. Albeit simple, this strategy is shown to facilitate the learning of discriminative features and thus improve event classification and localization performance. For example, after annotating the event location ground truths of the UCF-Crime test set, we report very competitive accuracy compared with the state of the art on both frame-level and pixel-level metrics, corresponding to classification and localization, respectively. [ABSTRACT FROM AUTHOR]

Missing values in data are common in real world applications. There are several methods that deal with this problem. In this paper we present lookahead selective sampling (LSS) algorithms for datasets with missing values. We developed two versions of selective sampling. The first one integrates a distance function that can measure the similarity between pairs of incomplete points within the framework of the LSS algorithm. The second algorithm uses ensemble clustering in order to represent the data in a cluster matrix without missing values and then run the LSS algorithm based on the ensemble clustering instance space (LSS-EC). To construct the cluster matrix, we use the k-means and mean shift clustering algorithms especially modified to deal with incomplete datasets. We tested our algorithms on six standard numerical datasets from different fields. On these datasets we simulated missing values and compared the performance of the LSS and LSS-EC algorithms for incomplete data to two other basic methods. Our experiments show that the suggested selective sampling algorithms outperform the other methods.

Modeling and predicting of a novel polar organic chemical integrative sampler (POCIS) for sampling of perfluorooctanoic acid (PFOA) and perfluorooctanesulfonic acid (PFOS) using molecularly imprinted polymers (MIPs) as receiving phase are presented in this study. Laboratory microcosm experiments were conducted to investigate the uptake kinetics, effects of flow velocity, pH, and dissolved organic matter (DOM), and also the selectivity of the POCIS. In this study, uptake study of PFOA and PFOS sampling on MIP-POCIS, over 14 days, was investigated. Laboratory calibrations of MIP-POCIS yielded sampling rate (Rs) values for PFOA and PFOS were 0.387 and 0.229 L/d, higher than POCIS using commercial sorbent WAX as receiving phase (0.133 and 0.141 L/d for PFOA and PFOS, respectively) in quiescent condition. The Rs values for PFOA and PFOS sampling on MIP-POCIS were increased to 0.591 and 0.281 L/d in stirred condition (0.01 m/s), and no significant increase occurred when the flow velocity was further increased. The Rs values were kept relatively high in the solution of which the pH was lower than the isoelectric point (IEP) of MIP-sorbent and decreased when the solution pH was extend the IEP value. Under the experimental conditions, DOM seemed to slightly facilitate the Rs values of PFOA and PFOS in MIP-POCIS. The results showed that the interaction between the target compounds and the receiving phase was fully integrated by the imprinting effects and also the electrostatic interaction. Finally, comparing the sampling rate of WAX-POCIS and the MIP-POCIS, the MIP-POCIS offers promising perspectives for selective sampling ability for PFOA and PFOS. [ABSTRACT FROM AUTHOR]

This paper analyzes the communication between two energy harvesting wireless sensor nodes. The nodes use automatic repeat request and forward error correction mechanism for the error control. The random nature of available energy and arrivals of harvested energy may induce interruption to the signal sampling and decoding operations. We propose a selective sampling scheme, where the length of the transmitted packet to be sampled depends on the available energy at the receiver. The receiver performs the decoding when complete samples of the packet are available. The selective sampling information bits are piggybacked on the automatic repeat request messages for the transmitter use. This way, the receiver node manages more efficiently its energy use. Besides, we present the partially observable Markov decision process formulation, which minimizes the long-term average pairwise error probability and optimizes the transmit power. Optimal and suboptimal power assignment strategies are introduced for retransmissions, which are adapted to the selective sampling and channel state information. With finite battery size and fixed power assignment policy, an analytical expression for the average packet drop probability is derived. Numerical simulations show the performance gain of the proposed scheme with power assignment strategy over the conventional scheme. [ABSTRACT FROM PUBLISHER]

In this paper, a robust and secure image hashing scheme based on salient structure features is proposed, which can be applied in image authentication and retrieval. In order to acquire the fixed length of image hash, the pre-processing for image regularization is first conducted on input image. Salient edge detection is then applied on the secondary image, and a series of non-overlapping blocks containing the richest structural information in the secondary image are selectively sampled according to the edge binary map. Dominant DCT coefficients of the sampled blocks with their corresponding position information are retrieved as the robust features. After the compression with dimensionality reduction for the concatenated features, the final hash can be produced. Experimental results show that the proposed scheme has better performances of perceptual robustness and discrimination compared with some state-of-the-art schemes. [ABSTRACT FROM AUTHOR]

Convolutional neural networks (CNNs) are deep learning network architectures that have pushed forward the state-of-the-art in a range of computer vision applications and are increasingly popular in medical image analysis. However, training of CNNs is time-consuming and challenging. In medical image analysis tasks, the majority of training examples are easy to classify and therefore contribute little to the CNN learning process. In this paper, we propose a method to improve and speed-up the CNN training for medical image analysis tasks by dynamically selecting misclassified negative samples during training. Training samples are heuristically sampled based on classification by the current status of the CNN. Weights are assigned to the training samples and informative samples are more likely to be included in the next CNN training iteration. We evaluated and compared our proposed method by training a CNN with (SeS) and without (NSeS) the selective sampling method. We focus on the detection of hemorrhages in color fundus images. A decreased training time from 170 epochs to 60 epochs with an increased performance—on par with two human experts—was achieved with areas under the receiver operating characteristics curve of 0.894 and 0.972 on two data sets. The SeS CNN statistically outperformed the NSeS CNN on an independent test set. [ABSTRACT FROM AUTHOR]

This work establishes distribution-free upper and lower bounds on the minimax label complexity of active learning with general hypothesis classes, under various noise models. The results reveal a number of surprising facts. In particular, under the noise model of Tsy-bakov (2004), the minimax label complexity of active learning with a VC class is always asymptotically smaller than that of passive learning, and is typically significantly smaller than the best previously-published upper bounds in the active learning literature. In high-noise regimes, it turns out that all active learning problems of a given VC dimension have roughly the same minimax label complexity, which contrasts with well-known results for bounded noise. In low-noise regimes, we find that the label complexity is well-characterized by a simple combinatorial complexity measure we call the star number. Interestingly, we find that almost all of the complexity measures previously explored in the active learning literature have worst-case values exactly equal to the star number. We also propose new active learning strategies that nearly achieve these minimax label complexities. [ABSTRACT FROM AUTHOR]

This paper brings up a very important issue for active learning in practice. Traditional active learning mechanism is based on the assumption that the number of classes happens to be known in advance, and thus selective sampling is confined to the determinate model. However, as is the case for many applications, the model class is usually indeterminate and there is every chance that the hypothesis itself is inappropriate. To address this problem, we propose a novel indeterminate multi-class active learning algorithm, which comprehensively evaluates the instance based on both the value in refining the existing model and the potential in triggering model rectification. In this way, balance is effectively achieved between model update and model upgrade. Advantage of the proposed algorithm is demonstrated by experiments of classification tasks on both synthetic and real-world dataset. [ABSTRACT FROM AUTHOR]

We introduce a new and improved characterization of the label complexity of disagreement-based active learning, in which the leading quantity is the version space compression set size. This quantity is defined as the size of the smallest subset of the training data that induces the same version space. We show various applications of the new characterization, including a tight analysis of CAL and refined label complexity bounds for linear separators under mixtures of Gaussians and axis-aligned rectangles under product densities. The version space compression set size, as well as the new characterization of the label complexity, can be naturally extended to agnostic learning problems, for which we show new speedup results for two well known active learning algorithms. [ABSTRACT FROM AUTHOR]

Gait analysis is an important process to gauge human motion. Recently, longitudinal gait analysis received much attention from the medical and healthcare domains. The challenge in studies over extended time periods is the battery life. Due to the continuous sensing and computing, wearable gait devices cannot fulfill a full-day work schedule. In this paper, we present an energy-efficient adaptive sensing framework to address this problem. Through presampling for content understanding, a selective sensing and sparsity-based signal reconstruction method is proposed. In particular, we develop and implement the new sensing scheme in a smart insole system to reduce the number of samples, while still preserving the information integrity of gait parameters. Experimental results show the effectiveness of our method in data point reduction. Our proposed method improves the battery life to 10.47 h, while normalized mean square error is within 10%. [ABSTRACT FROM AUTHOR]

Benefits of digital current mode control are often limited by the choice of a current-loop sampling rate. A higher rate requires a fast analog-to-digital converter that consumes substantial power and increases cost. A lower rate often results in subharmonic oscillations, even using a programmable ramp compensation. This brief proposes a simple technique to compute the steady-state duty ratio in real time, when the closed-loop controller is in action. A time-to-digital converter translates cycle-by-cycle duty ratio information into digital code, and a “duty ratio computation” block generates the computed duty ratio using a moving average filter. At steady state, this enforces a virtual open-loop configuration and completely disables current-loop sampling and controller computation, thereby saving substantial power and eliminating subharmonic oscillations. Considering a dc–dc buck converter as the test case, it is found that even in the presence of high periodic behavior under the closed-loop control, a near-ideal steady-state duty ratio can be reconstructed. Design-related issues along with duty ratio saturation are discussed with test cases. [ABSTRACT FROM PUBLISHER]