scispace - formally typeset
Search or ask a question

Showing papers on "Matching (statistics) published in 2017"


Proceedings ArticleDOI
13 Feb 2017
TL;DR: This article proposed a bilateral multi-perspective matching (BiMPM) model under the "matching-aggregation" framework, which first encodes two sentences with a BiLSTM encoder and then matches the two encoded sentences in two directions.
Abstract: Natural language sentence matching is a fundamental technology for a variety of tasks. Previous approaches either match sentences from a single direction or only apply single granular (word-by-word or sentence-by-sentence) matching. In this work, we propose a bilateral multi-perspective matching (BiMPM) model under the "matching-aggregation" framework. Given two sentences $P$ and $Q$, our model first encodes them with a BiLSTM encoder. Next, we match the two encoded sentences in two directions $P \rightarrow Q$ and $P \leftarrow Q$. In each matching direction, each time step of one sentence is matched against all time-steps of the other sentence from multiple perspectives. Then, another BiLSTM layer is utilized to aggregate the matching results into a fix-length matching vector. Finally, based on the matching vector, the decision is made through a fully connected layer. We evaluate our model on three tasks: paraphrase identification, natural language inference and answer sentence selection. Experimental results on standard benchmark datasets show that our model achieves the state-of-the-art performance on all tasks.

563 citations


Posted Content
01 Jan 2017
TL;DR: This article developed a non-parametric causal forest for estimating heterogeneous treatment effects that extends Breiman's widely used random forest algorithm and showed that causal forests are pointwise consistent for the true treatment effect, and have an asymptotic Gaussian and centered sampling distribution.
Abstract: Many scientific and engineering challenges--ranging from personalized medicine to customized marketing recommendations--require an understanding of treatment effect heterogeneity. In this paper, we develop a non-parametric causal forest for estimating heterogeneous treatment effects that extends Breiman's widely used random forest algorithm. In the potential outcomes framework with unconfoundedness, we show that causal forests are pointwise consistent for the true treatment effect, and have an asymptotically Gaussian and centered sampling distribution. We also discuss a practical method for constructing asymptotic confidence intervals for the true treatment effect that are centered at the causal forest estimates. Our theoretical results rely on a generic Gaussian theory for a large family of random forest algorithms. To our knowledge, this is the first set of results that allows any type of random forest, including classification and regression forests, to be used for provably valid statistical inference. In experiments, we find causal forests to be substantially more powerful than classical methods based on nearest-neighbor matching, especially in the presence of irrelevant covariates.

485 citations


Proceedings ArticleDOI
Yu Wu1, Wei Wu2, Chen Xing3, Ming Zhou2, Zhoujun Li1 
05 Aug 2017
TL;DR: Zhang et al. as mentioned in this paper proposed a sequential matching network (SMN) which first matches a response with each utterance in the context on multiple levels of granularity, and distills important matching information from each pair as a vector with convolution and pooling operations.
Abstract: We study response selection for multi-turn conversation in retrieval based chatbots. Existing work either concatenates utterances in context or matches a response with a highly abstract context vector finally, which may lose relationships among the utterances or important information in the context. We propose a sequential matching network (SMN) to address both problems. SMN first matches a response with each utterance in the context on multiple levels of granularity, and distills important matching information from each pair as a vector with convolution and pooling operations. The vectors are then accumulated in a chronological order through a recurrent neural network (RNN) which models relationships among the utterances. The final matching score is calculated with the hidden states of the RNN. Empirical study on two public data sets shows that SMN can significantly outperform state-of-the-art methods for response selection in multi-turn conversation.

484 citations


Proceedings ArticleDOI
TL;DR: Deep Relevance Matching (DRMM) as mentioned in this paper employs a joint deep architecture at the query term level for relevance matching, using matching histogram mapping, a feed forward matching network, and a term gating network.
Abstract: In recent years, deep neural networks have led to exciting breakthroughs in speech recognition, computer vision, and natural language processing (NLP) tasks. However, there have been few positive results of deep models on ad-hoc retrieval tasks. This is partially due to the fact that many important characteristics of the ad-hoc retrieval task have not been well addressed in deep models yet. Typically, the ad-hoc retrieval task is formalized as a matching problem between two pieces of text in existing work using deep models, and treated equivalent to many NLP tasks such as paraphrase identification, question answering and automatic conversation. However, we argue that the ad-hoc retrieval task is mainly about relevance matching while most NLP matching tasks concern semantic matching, and there are some fundamental differences between these two matching tasks. Successful relevance matching requires proper handling of the exact matching signals, query term importance, and diverse matching requirements. In this paper, we propose a novel deep relevance matching model (DRMM) for ad-hoc retrieval. Specifically, our model employs a joint deep architecture at the query term level for relevance matching. By using matching histogram mapping, a feed forward matching network, and a term gating network, we can effectively deal with the three relevance matching factors mentioned above. Experimental results on two representative benchmark collections show that our model can significantly outperform some well-known retrieval models as well as state-of-the-art deep matching models.

477 citations


Posted Content
TL;DR: This work proposes a bilateral multi-perspective matching (BiMPM) model under the "matching-aggregation" framework that achieves the state-of-the-art performance on all tasks.
Abstract: Natural language sentence matching is a fundamental technology for a variety of tasks. Previous approaches either match sentences from a single direction or only apply single granular (word-by-word or sentence-by-sentence) matching. In this work, we propose a bilateral multi-perspective matching (BiMPM) model under the "matching-aggregation" framework. Given two sentences $P$ and $Q$, our model first encodes them with a BiLSTM encoder. Next, we match the two encoded sentences in two directions $P \rightarrow Q$ and $P \leftarrow Q$. In each matching direction, each time step of one sentence is matched against all time-steps of the other sentence from multiple perspectives. Then, another BiLSTM layer is utilized to aggregate the matching results into a fix-length matching vector. Finally, based on the matching vector, the decision is made through a fully connected layer. We evaluate our model on three tasks: paraphrase identification, natural language inference and answer sentence selection. Experimental results on standard benchmark datasets show that our model achieves the state-of-the-art performance on all tasks.

427 citations


Posted Content
TL;DR: This paper proposes a novel method called AlignedReID that extracts a global feature which is jointly learned with local features, and is the first to surpass human-level performance on Market1501 and CUHK03, two widely used Person ReID datasets.
Abstract: In this paper, we propose a novel method called AlignedReID that extracts a global feature which is jointly learned with local features. Global feature learning benefits greatly from local feature learning, which performs an alignment/matching by calculating the shortest path between two sets of local features, without requiring extra supervision. After the joint learning, we only keep the global feature to compute the similarities between images. Our method achieves rank-1 accuracy of 94.4% on Market1501 and 97.8% on CUHK03, outperforming state-of-the-art methods by a large margin. We also evaluate human-level performance and demonstrate that our method is the first to surpass human-level performance on Market1501 and CUHK03, two widely used Person ReID datasets.

426 citations


Journal ArticleDOI
TL;DR: In this article, the authors compared the performance of conventional covariate adjustment with four common PS methods: matching, stratification, inverse probability weighting, and use of PS as a covariate.

413 citations


Proceedings ArticleDOI
01 Jul 2017
TL;DR: An extensive experimental evaluation of learned local features to establish a single evaluation protocol that ensures comparable results in terms of matching performance and describes the different descriptors regarding standard criteria.
Abstract: Matching local image descriptors is a key step in many computer vision applications. For more than a decade, hand-crafted descriptors such as SIFT have been used for this task. Recently, multiple new descriptors learned from data have been proposed and shown to improve on SIFT in terms of discriminative power. This paper is dedicated to an extensive experimental evaluation of learned local features to establish a single evaluation protocol that ensures comparable results. In terms of matching performance, we evaluate the different descriptors regarding standard criteria. However, considering matching performance in isolation only provides an incomplete measure of a descriptors quality. For example, finding additional correct matches between similar images does not necessarily lead to a better performance when trying to match images under extreme viewpoint or illumination changes. Besides pure descriptor matching, we thus also evaluate the different descriptors in the context of image-based reconstruction. This enables us to study the descriptor performance on a set of more practical criteria including image retrieval, the ability to register images under strong viewpoint and illumination changes, and the accuracy and completeness of the reconstructed cameras and scenes. To facilitate future research, the full evaluation pipeline is made publicly available.

306 citations


Posted Content
TL;DR: This paper compares the performance of three different image matching techniques, i.e., SIFT, SURF, and ORB, against different kinds of transformations and deformations such as scaling, rotation, noise, fish eye distortion, and shearing and shows that which algorithm is the best more robust against each kind of distortion.
Abstract: Fast and robust image matching is a very important task with various applications in computer vision and robotics. In this paper, we compare the performance of three different image matching techniques, i.e., SIFT, SURF, and ORB, against different kinds of transformations and deformations such as scaling, rotation, noise, fish eye distortion, and shearing. For this purpose, we manually apply different types of transformations on original images and compute the matching evaluation parameters such as the number of key points in images, the matching rate, and the execution time required for each algorithm and we will show that which algorithm is the best more robust against each kind of distortion. Index Terms-Image matching, scale invariant feature transform (SIFT), speed up robust feature (SURF), robust independent elementary features (BRIEF), oriented FAST, rotated BRIEF (ORB).

261 citations


Journal ArticleDOI
TL;DR: It is shown that regression adjustment could dramatically remove residual confounding bias when it included all of the covariates with a standardized difference greater than 0.10.
Abstract: Double-adjustment can be used to remove confounding if imbalance exists after propensity score (PS) matching. However, it is not always possible to include all covariates in adjustment. We aimed to find the optimal imbalance threshold for entering covariates into regression. We conducted a series of Monte Carlo simulations on virtual populations of 5,000 subjects. We performed PS 1:1 nearest-neighbor matching on each sample. We calculated standardized mean differences across groups to detect any remaining imbalance in the matched samples. We examined 25 thresholds (from 0.01 to 0.25, stepwise 0.01) for considering residual imbalance. The treatment effect was estimated using logistic regression that contained only those covariates considered to be unbalanced by these thresholds. We showed that regression adjustment could dramatically remove residual confounding bias when it included all of the covariates with a standardized difference greater than 0.10. The additional benefit was negligible when we also adjusted for covariates with less imbalance. We found that the mean squared error of the estimates was minimized under the same conditions. If covariate balance is not achieved, we recommend reiterating PS modeling until standardized differences below 0.10 are achieved on most covariates. In case of remaining imbalance, a double adjustment might be worth considering.

258 citations


Posted Content
TL;DR: The model aims to learn an asymmetric metric, i.e., specific projection for each view, based on asymmetric clustering on cross-view person images, and finds a shared space where view-specific bias is alleviated and thus better matching performance can be achieved.
Abstract: While metric learning is important for Person re-identification (RE-ID), a significant problem in visual surveillance for cross-view pedestrian matching, existing metric models for RE-ID are mostly based on supervised learning that requires quantities of labeled samples in all pairs of camera views for training. However, this limits their scalabilities to realistic applications, in which a large amount of data over multiple disjoint camera views is available but not labelled. To overcome the problem, we propose unsupervised asymmetric metric learning for unsupervised RE-ID. Our model aims to learn an asymmetric metric, i.e., specific projection for each view, based on asymmetric clustering on cross-view person images. Our model finds a shared space where view-specific bias is alleviated and thus better matching performance can be achieved. Extensive experiments have been conducted on a baseline and five large-scale RE-ID datasets to demonstrate the effectiveness of the proposed model. Through the comparison, we show that our model works much more suitable for unsupervised RE-ID compared to classical unsupervised metric learning models. We also compare with existing unsupervised RE-ID methods, and our model outperforms them with notable margins. Specifically, we report the results on large-scale unlabelled RE-ID dataset, which is important but unfortunately less concerned in literatures.

Journal ArticleDOI
TL;DR: This paper employs the Gale–Shapley algorithm to match D2D pairs with cellular UEs, which is proved to be stable and weak Pareto optimal, and extends the algorithm to address scalability issues in large-scale networks by developing tie-breaking and preference-deletion-based matching rules.
Abstract: Energy-efficiency (EE) is critical for device-to-device (D2D) enabled cellular networks due to limited battery capacity and severe cochannel interference. In this paper, we address the EE optimization problem by adopting a stable matching approach. The NP-hard joint resource allocation problem is formulated as a one-to-one matching problem under two-sided preferences, which vary dynamically with channel states and interference levels. A game-theoretic approach is employed to analyze the interactions and correlations among user equipments (UEs), and an iterative power allocation algorithm is developed to establish mutual preferences based on nonlinear fractional programing. We then employ the Gale–Shapley algorithm to match D2D pairs with cellular UEs, which is proved to be stable and weak Pareto optimal. We provide a theoretical analysis and description for implementation details and algorithmic complexity. We also extend the algorithm to address scalability issues in large-scale networks by developing tie-breaking and preference-deletion-based matching rules. Simulation results validate the theoretical analysis and demonstrate that significant performance gains of average EE and matching satisfactions can be achieved by the proposed algorithm.

Proceedings Article
05 Sep 2017
TL;DR: In this paper, the authors investigate the non-identifiability issues associated with bidirectional adversarial training for joint distribution matching and propose both adversarial and non-adversarial approaches to learn desirable matched joint distributions for unsupervised and supervised tasks.
Abstract: We investigate the non-identifiability issues associated with bidirectional adversarial training for joint distribution matching. Within a framework of conditional entropy, we propose both adversarial and non-adversarial approaches to learn desirable matched joint distributions for unsupervised and supervised tasks. We unify a broad family of adversarial models as joint distribution matching problems. Our approach stabilizes learning of unsupervised bidirectional adversarial learning methods. Further, we introduce an extension for semi-supervised learning tasks. Theoretical results are validated in synthetic data and real-world applications.

Journal ArticleDOI
TL;DR: Two approaches for estimating the Average Treatment Effect (ATE) on survival outcomes: Inverse Probability of Treatment Weighting (IPTW) and full matching are compared in an extensive set of simulations that varied the extent of confounding and the amount of misspecification of the propensity score model.
Abstract: There is increasing interest in estimating the causal effects of treatments using observational data Propensity-score matching methods are frequently used to adjust for differences in observed characteristics between treated and control individuals in observational studies Survival or time-to-event outcomes occur frequently in the medical literature, but the use of propensity score methods in survival analysis has not been thoroughly investigated This paper compares two approaches for estimating the Average Treatment Effect (ATE) on survival outcomes: Inverse Probability of Treatment Weighting (IPTW) and full matching The performance of these methods was compared in an extensive set of simulations that varied the extent of confounding and the amount of misspecification of the propensity score model We found that both IPTW and full matching resulted in estimation of marginal hazard ratios with negligible bias when the ATE was the target estimand and the treatment-selection process was weak to moderate However, when the treatment-selection process was strong, both methods resulted in biased estimation of the true marginal hazard ratio, even when the propensity score model was correctly specified When the propensity score model was correctly specified, bias tended to be lower for full matching than for IPTW The reasons for these biases and for the differences between the two methods appeared to be due to some extreme weights generated for each method Both methods tended to produce more extreme weights as the magnitude of the effects of covariates on treatment selection increased Furthermore, more extreme weights were observed for IPTW than for full matching However, the poorer performance of both methods in the presence of a strong treatment-selection process was mitigated by the use of IPTW with restriction and full matching with a caliper restriction when the propensity score model was correctly specified

Proceedings ArticleDOI
21 Jul 2017
TL;DR: Zhang et al. as mentioned in this paper proposed a selective multimodal Long Short-Term Memory network (sm-LSTM) for instance-aware image and sentence matching based on the observation that such a global similarity arises from a complex aggregation of multiple local similarities between pairwise instances of image (objects) and sentence (words).
Abstract: Effective image and sentence matching depends on how to well measure their global visual-semantic similarity. Based on the observation that such a global similarity arises from a complex aggregation of multiple local similarities between pairwise instances of image (objects) and sentence (words), we propose a selective multimodal Long Short-Term Memory network (sm-LSTM) for instance-aware image and sentence matching. The sm-LSTM includes a multimodal context-modulated attention scheme at each timestep that can selectively attend to a pair of instances of image and sentence, by predicting pairwise instance-aware saliency maps for image and sentence. For selected pairwise instances, their representations are obtained based on the predicted saliency maps, and then compared to measure their local similarity. By similarly measuring multiple local similarities within a few timesteps, the sm-LSTM sequentially aggregates them with hidden states to obtain a final matching score as the desired global similarity. Extensive experiments show that our model can well match image and sentence with complex content, and achieve the state-of-the-art results on two public benchmark datasets.

Journal ArticleDOI
TL;DR: In this paper, price matching is used as a short-term strategy to counter show-rooming behavior in brick-and-mortar stores, which is referred to as free-riding behavior by customers.
Abstract: Customers often evaluate products at brick-and-mortar stores to identify their “best-fit” product but buy it for a lower price at a competing online retailer. This free-riding behavior by customers is referred to as “showrooming,” and we show that this is detrimental to the profits of the brick-and-mortar stores. We first analyze price matching as a short-term strategy to counter showrooming. Price matching allows customers to purchase a product from the store for less than the store’s posted price, so one would expect the price matching strategy to be less effective as the fraction of customers who seek the matching increases. However, our results show that with an increase in the fraction of customers who seek price matching, the store’s profits initially decrease and then increase. While price matching could be used even when customers do not exhibit showrooming behavior, we find that it is more effective when customers do showrooming. We then study exclusivity of product assortments as a long-term str...

Journal ArticleDOI
TL;DR: Zhang et al. as mentioned in this paper proposed a novel perspective for person re-identification based on learning person saliency and matching saliency distribution, which consists of four steps: to handle misalignment caused by drastic viewpoint change and pose variations, they apply adjacency constrained patch matching to build dense correspondence between image pairs.
Abstract: Human eyes can recognize person identities based on small salient regions, i.e., person saliency is distinctive and reliable in pedestrian matching across disjoint camera views. However, such valuable information is often hidden when computing similarities of pedestrian images with existing approaches. Inspired by our user study result of human perception on person saliency, we propose a novel perspective for person re-identification based on learning person saliency and matching saliency distribution. The proposed saliency learning and matching framework consists of four steps: (1) To handle misalignment caused by drastic viewpoint change and pose variations, we apply adjacency constrained patch matching to build dense correspondence between image pairs. (2) We propose two alternative methods, i.e., K-Nearest Neighbors and One-class SVM, to estimate a saliency score for each image patch, through which distinctive features stand out without using identity labels in the training procedure. (3) saliency matching is proposed based on patch matching. Matching patches with inconsistent saliency brings penalty, and images of the same identity are recognized by minimizing the saliency matching cost. (4) Furthermore, saliency matching is tightly integrated with patch matching in a unified structural RankSVM learning framework. The effectiveness of our approach is validated on the four public datasets. Our approach outperforms the state-of-the-art person re-identification methods on all these datasets.

Proceedings ArticleDOI
01 Oct 2017
TL;DR: A novel global ranking algorithm, applied to a Markov network built upon the 3D map, which takes account of not only visual similarities between individual 2D-3D matches, but also their global compatibilities among all matching pairs found in the scene.
Abstract: Given an image of a street scene in a city, this paper develops a new method that can quickly and precisely pinpoint at which location (as well as viewing direction) the image was taken, against a pre-stored large-scale 3D point-cloud map of the city. We adopt the recently developed 2D-3D direct feature matching framework for this task [23,31,32,42–44]. This is a challenging task especially for large-scale problems. As the map size grows bigger, many 3D points in the wider geographical area can be visually very similar–or even identical–causing severe ambiguities in 2D-3D feature matching. The key is to quickly and unambiguously find the correct matches between a query image and the large 3D map. Existing methods solve this problem mainly via comparing individual features’ visual similarities in a local and per feature manner, thus only local solutions can be found, inadequate for large-scale applications. In this paper, we introduce a global method which harnesses global contextual information exhibited both within the query image and among all the 3D points in the map. This is achieved by a novel global ranking algorithm, applied to a Markov network built upon the 3D map, which takes account of not only visual similarities between individual 2D-3D matches, but also their global compatibilities (as measured by co-visibility) among all matching pairs found in the scene. Tests on standard benchmark datasets show that our method achieved both higher precision and comparable recall, compared with the state-of-the-art.

Proceedings ArticleDOI
01 Jul 2017
TL;DR: In this paper, a bi-directional neural network architecture for matching vectors from two data sources is introduced, which employs two tied neural network channels that project the two views into a common, maximally correlated space using the Euclidean loss.
Abstract: Linking two data sources is a basic building block in numerous computer vision problems. Canonical Correlation Analysis (CCA) achieves this by utilizing a linear optimizer in order to maximize the correlation between the two views. Recent work makes use of non-linear models, including deep learning techniques, that optimize the CCA loss in some feature space. In this paper, we introduce a novel, bi-directional neural network architecture for the task of matching vectors from two data sources. Our approach employs two tied neural network channels that project the two views into a common, maximally correlated space using the Euclidean loss. We show a direct link between the correlation-based loss and Euclidean loss, enabling the use of Euclidean loss for correlation maximization. To overcome common Euclidean regression optimization problems, we modify well-known techniques to our problem, including batch normalization and dropout. We show state of the art results on a number of computer vision matching tasks including MNIST image matching and sentence-image matching on the Flickr8k, Flickr30k and COCO datasets.

Proceedings ArticleDOI
01 Jul 2017
TL;DR: The proposed pipeline achieves state of the art accuracy on the largest and most competitive stereo benchmarks, and the learned confidence is shown to outperform all existing alternatives.
Abstract: We present an improved three-step pipeline for the stereo matching problem and introduce multiple novelties at each stage. We propose a new highway network architecture for computing the matching cost at each possible disparity, based on multilevel weighted residual shortcuts, trained with a hybrid loss that supports multilevel comparison of image patches. A novel post-processing step is then introduced, which employs a second deep convolutional neural network for pooling global information from multiple disparities. This network outputs both the image disparity map, which replaces the conventional winner takes all strategy, and a confidence in the prediction. The confidence score is achieved by training the network with a new technique that we call the reflective loss. Lastly, the learned confidence is employed in order to better detect outliers in the refinement step. The proposed pipeline achieves state of the art accuracy on the largest and most competitive stereo benchmarks, and the learned confidence is shown to outperform all existing alternatives.

Journal ArticleDOI
TL;DR: In this paper, the authors consider a notion of stability for ride-share matches and present several mathematical programming methods to establish stable or nearly stable matches, where they note that ride-sharing matching optimization is performed over time with incomplete information.
Abstract: Dynamic ride-sharing systems enable people to share rides and increase the efficiency of urban transportation by connecting riders and drivers on short notice. Automated systems that establish ride-share matches with minimal input from participants provide convenience and the most potential for system-wide performance improvement, such as reduction in total vehicle-miles traveled. Indeed, such systems may be designed to match riders and drivers to maximize system performance improvement. However, system-optimal matches may not provide the maximum benefit to each individual participant. In this paper, we consider a notion of stability for ride-share matches and present several mathematical programming methods to establish stable or nearly stable matches, where we note that ride-share matching optimization is performed over time with incomplete information. Our numerical experiments using travel demand data for the metropolitan Atlanta region show that we can significantly increase the stability of ride-sha...

Journal ArticleDOI
TL;DR: In this paper, the authors study the competition in matching markets with random heterogeneous preferences and an unequal number of agents on either side, and show that even the slightest imbalance yields an essentially unique stable matching.
Abstract: We study competition in matching markets with random heterogeneous preferences and an unequal number of agents on either side. First, we show that even the slightest imbalance yields an essentially unique stable matching. Second, we give a tight description of stable outcomes, showing that matching markets are extremely competitive. Each agent on the short side of the market is matched with one of his top choices, and each agent on the long side either is unmatched or does almost no better than being matched with a random partner. Our results suggest that any matching market is likely to have a small core, explaining why small cores are empirically ubiquitous.

Journal ArticleDOI
TL;DR: In this article, the authors explore the nontransferable and perfectly transferable utility matching paradigms, and then a unifying imperfectly transferable matching model for assortative matching.
Abstract: Toward understanding assortative matching, this is a self-contained introduction to research on search and matching. We first explore the nontransferable and perfectly transferable utility matching paradigms, and then a unifying imperfectly transferable utility matching model. Motivated by some unrealistic predictions of frictionless matching, we flesh out the foundational economics of search theory. We then revisit the original matching paradigms with search frictions. We finally allow informational frictions that often arise, such as in college-student sorting.

Proceedings ArticleDOI
01 Oct 2017
TL;DR: This paper presents a framework for learning stereo matching costs without human supervision by updating network parameters in an iterative manner and performs even comparably with other supervised methods.
Abstract: Convolutional neural networks showed the ability in stereo matching cost learning. Recent approaches learned parameters from public datasets that have ground truth disparity maps. Due to the difficulty of labeling ground truth depth, usable data for system training is rather limited, making it difficult to apply the system to real applications. In this paper, we present a framework for learning stereo matching costs without human supervision. Our method updates network parameters in an iterative manner. It starts with a randomly initialized network. Left-right check is adopted to guide the training. Suitable matching is then picked and used as training data in following iterations. Our system finally converges to a stable state and performs even comparably with other supervised methods.

Posted Content
TL;DR: This paper investigates two-branch neural networks for learning the similarity between image-sentence matching and region-phrase matching, and proposes two network structures that produce different output representations.
Abstract: Image-language matching tasks have recently attracted a lot of attention in the computer vision field. These tasks include image-sentence matching, i.e., given an image query, retrieving relevant sentences and vice versa, and region-phrase matching or visual grounding, i.e., matching a phrase to relevant regions. This paper investigates two-branch neural networks for learning the similarity between these two data modalities. We propose two network structures that produce different output representations. The first one, referred to as an embedding network, learns an explicit shared latent embedding space with a maximum-margin ranking loss and novel neighborhood constraints. Compared to standard triplet sampling, we perform improved neighborhood sampling that takes neighborhood information into consideration while constructing mini-batches. The second network structure, referred to as a similarity network, fuses the two branches via element-wise product and is trained with regression loss to directly predict a similarity score. Extensive experiments show that our networks achieve high accuracies for phrase localization on the Flickr30K Entities dataset and for bi-directional image-sentence retrieval on Flickr30K and MSCOCO datasets.

Journal ArticleDOI
TL;DR: The authors develop strategies to estimate false discovery rates (FDR) by empirical Bayes and target-decoy based methods which enable a user to define the scoring criteria for spectral matching.
Abstract: The annotation of small molecules in untargeted mass spectrometry relies on the matching of fragment spectra to reference library spectra. While various spectrum-spectrum match scores exist, the field lacks statistical methods for estimating the false discovery rates (FDR) of these annotations. We present empirical Bayes and target-decoy based methods to estimate the false discovery rate (FDR) for 70 public metabolomics data sets. We show that the spectral matching settings need to be adjusted for each project. By adjusting the scoring parameters and thresholds, the number of annotations rose, on average, by +139% (ranging from −92 up to +5705%) when compared with a default parameter set available at GNPS. The FDR estimation methods presented will enable a user to assess the scoring criteria for large scale analysis of mass spectrometry based metabolomics data that has been essential in the advancement of proteomics, transcriptomics, and genomics science. Matching fragment spectra to reference library spectra is an important procedure for annotating small molecules in untargeted mass spectrometry based metabolomics studies. Here, the authors develop strategies to estimate false discovery rates (FDR) by empirical Bayes and target-decoy based methods which enable a user to define the scoring criteria for spectral matching.

Journal ArticleDOI
TL;DR: In this paper, a learning framework based on a problem-specific Markov chain is proposed for underlay D2D network, where a two phase algorithm is developed to perform mode selection and resource allocation in the respective phases.
Abstract: Device to device (D2D) communication is considered as an effective technology for enhancing the spectral efficiency and network throughput of existing cellular networks. However, enabling it in an underlay fashion poses a significant challenge pertaining to interference management. In this paper, mode selection and resource allocation for an underlay D2D network is studied while simultaneously providing interference management. The problem is formulated as a combinatorial optimization problem whose objective is to maximize the utility of all D2D pairs. To solve this problem, a learning framework is proposed based on a problem-specific Markov chain. From the local balance equation of the designed Markov chain, the transition probabilities are derived for distributed implementation. Then, a novel two phase algorithm is developed to perform mode selection and resource allocation in the respective phases. This algorithm is then shown to converge to a near optimal solution. Moreover, to reduce the computation in the learning framework, two resource allocation algorithms based on matching theory are proposed to output a specific and deterministic solution. The first algorithm employs the one-to-one matching game approach whereas in the second algorithm, the one-to many matching game with externalities and dynamic quota is employed. Simulation results show that the proposed framework converges to a near optimal solution under all scenarios with probability one. Moreover, our results show that the proposed matching game with externalities achieves a performance gain of up to 35 percent in terms of the average utility compared to a classical matching scheme with no externalities.

Journal ArticleDOI
TL;DR: The concept and model of supply-demand matching hypernetwork ( Matching_Net) of manufacturing services in SOM system are put forward considering both manufacturing services and manufacturing tasks and the simulation is carried out to validate and verify the proposed models and the corresponding modeling method.
Abstract: The supply-demand matching and optimal-allocation of manufacturing resource and manufacturing capability, is always one of the key scientific issues to be addressed and the common objectives to be pursued for various advanced manufacturing systems (AMSs). Especially to the service-oriented manufacturing (SOM) systems (e.g., cloud manufacturing), which is a kind of typical and hot AMSs nowadays , it is extremely difficult to achieve the optimal allocation of large scales of different manufacturing services and the collaboration of manufacturing activities and business among massive manufacturing enterprises (no matter manufacturing service providers and consumers). Thus, under the current environment of social competition and collaboration, the most important challenge is how to integrate a variety of distributed manufacturing resources and capabilities in the form of manufacturing services efficiently and cost-effectively, as well as various demands or manufacturing tasks. In response to this challenge, the concept of supply-demand matching hypernetwork ( Matching_Net ) of manufacturing services in SOM system is put forward firstly in this paper. The proposed Matching_Net is constructed with manufacturing service network ( S_Net ), manufacturing task network ( T_Net ), and hyper-edges between those two networks which are revealing the matchable correlations between each service (supply) and each task (demand). Secondly, by comparing with the input or output information of manufacturing services and tasks from the functional view, the intelligent modeling of Matching_Net is illustrated and divided into those three constituent parts respectively. Finally, the simulation is carried out to validate and verify the proposed models and the modeling method. The model of hypernetwork (i.e., network of networks) is introduced into the supply-demand matching issues in service-oriented manufacturing systems.The concept and model of supply-demand matching hypernetwork ( Matching_Net ) of manufacturing services in SOM system are put forward considering both manufacturing services and manufacturing tasks.By comparing with the input or output information of manufacturing services and tasks, the intelligent modeling flow of Matching_Net is proposed and illustrated.The simulation is carried out to validate and verify the proposed models and the corresponding modeling method.

Journal ArticleDOI
TL;DR: A new deep coupled metric learning method for cross-modal matching, which aims to match samples captured from two different modalities into a shared latent feature subspace, under which the intraclass variation is minimized and the interclass variation is maximized.
Abstract: In this paper, we propose a new deep coupled metric learning (DCML) method for cross-modal matching, which aims to match samples captured from two different modalities (e.g., texts versus images, visible versus near infrared images). Unlike existing cross-modal matching methods which learn a linear common space to reduce the modality gap, our DCML designs two feedforward neural networks which learn two sets of hierarchical nonlinear transformations (one set for each modality) to nonlinearly map samples from different modalities into a shared latent feature subspace, under which the intraclass variation is minimized and the interclass variation is maximized, and the difference of each data pair captured from two modalities of the same class is minimized, respectively. Experimental results on four different cross-modal matching datasets validate the efficacy of the proposed approach.

Journal ArticleDOI
TL;DR: In this article, the manipulability of stable matching mechanisms is studied in markets with random cardinal utilities, which induce ordinal preferences over match partners, and it is shown that most agents in large matching markets are close to being indifferent of overall stable matchings.
Abstract: We study the manipulability of stable matching mechanisms. To quantify incentives to manipulate stable mechanisms, we consider markets with random cardinal utilities, which induce ordinal preferences over match partners. We show that most agents in large matching markets are close to being indifferent of overall stable matchings. In one-to-one matching, the utility gain by manipulating a stable mechanism does not exceed the gap between utilities from the best and worst stable partners. Thus, most agents in a large market would not have significant incentives to manipulate stable mechanisms. The incentive compatibility extends to many-to-one matching when agents employ truncation strategies and capacity manipulations in a Gale—Shapley mechanism.