Search

Mobile crowd sensing (MCS) is a popular sensing paradigm that leverages the power of massive mobile workers to perform various location-based sensing tasks. To assign workers with suitable tasks, recent research works investigated mobility prediction methods based on probabilistic and statistical models to estimate the worker's moving behavior, based on which the allocation algorithm is designed to match workers with tasks such that workers do not need to deviate from their daily routes and tasks can be completed as many as possible. In this paper, we propose a new multi-task allocation method based on mobility prediction, which differs from the existing works by (1) making use of workers' historical trajectories more comprehensively by using the fuzzy logic system to obtain more accurate mobility prediction and (2) designing a global heuristic searching algorithm to optimize the overall task completion rate based on the mobility prediction result, which jointly considers workers' and tasks' spatiotemporal features. We evaluate the proposed prediction method and task allocation algorithm using two real-world datasets. The experimental results validate the effectiveness of the proposed methods compared against baselines.

This paper aims to solve two important issues that frequently occur in existing automatic personality analysis systems: 1. Attempting to use very short video segments or even single frames to infer personality traits; 2. Lack of methods to encode person-specific facial dynamics for personality recognition. Hence, we proposes a novel Rank Loss which utilizes the natural temporal evolution of facial actions, rather than personality labels, for self-supervised learning of facial dynamics. Our approach first trains a generic U-net model that can infer general facial dynamics learned from unlabelled face videos. Then, the generic model is frozen, and a set of intermediate filters are incorporated into this architecture. The self-supervised learning is then resumed with only person-specific videos. This way, the learned filters' weights are person-specific, making them a valuable source for modeling person-specific facial dynamics. We then concatenate the weights of the learned filters as a person-specific representation, which can be directly used to predict the personality traits without needing other parts of the network. We evaluate the proposed approach on both self-reported personality and apparent personality datasets. Besides achieving promising results in personality trait estimation from videos, we show that fusion of tasks reaches highest accuracy, and that multi-scale dynamics are more informative than single-scale dynamics.

The focal role of flow in promoting desirable outcomes in companies, such as increased employees' well-being and performance, led scholars to study flow in the context of work. However, current measurement approaches which assess flow via self-report scales after task execution are limited due to obtrusiveness and a lack of real-time support. Hence, new measurement approaches must be created to overcome these limitations. In this paper, we use cardiac features (heart rate variability; HRV) and a Random Forest classifier to distinguish high and low flow. Our results from a large-scale lab experiment with 158 participants and a field study with nine participants reveal, that with HRV features alone, flow-classifiers can be built with an accuracy of 68.5% (lab) and 70.6% (field). Our research contributes to the challenge of developing a less obtrusive, real-time measurement method of flow based on physiological features and to investigate flow from a physiological perspective. Our findings may serve as foundation for future work aiming to build physio-adaptive systems which can improve employee's performance. For instance, these systems could ensure that no notifications are forwarded to employees while the system is 'sensing' flow.

The analysis of variance in complex text traditions is an arduous task when carried out manually. Text alignment algorithms provide domain experts with a robust alternative to such repetitive tasks. Existing white-box approaches allow the digital humanities to establish syntax-based metrics taking into account the spelling, morphology and order of words. However, they produce limited results, as semantic meanings are typically not taken into account. Our interdisciplinary collaboration between visualization and digital humanities combined a semi-supervised text alignment approach based on word embeddings that take not only syntactic but also semantic text features into account, thereby improving the overall quality of the alignment. In our collaboration, we developed different visual interfaces that communicate the word distribution in high-dimensional vector space generated by the underlying neural network for increased transparency, assessment of the tool's reliability and overall improved hypothesis generation. We further offer visual means to enable the expert reader to feed domain knowledge into the system at multiple levels with the aim of improving both the product and the process of text alignment. This ultimately illustrates how visualization can engage with and augment complex modes of reading in the humanities.

In this work, we study an important task in location-based services, namely Personalized Route Recommendation (PRR). Given a road network, the PRR task aims to generate user-specific route suggestions for replying to users' route queries. A classic approach is to adapting search algorithms to construct pathfinding-like solutions. These methods typically focus on reducing search space with suitable heuristic strategies. For these search algorithms, heuristic strategies are often handcrafted, which are not flexible to work in complicated task settings. In this paper, we propose to improve search algorithms with neural networks for solving the PRR task based on the widely used A* algorithm. Our model consists of two components. First, we employ attention-based Recurrent Neural Networks (RNN) to model the cost from the source to the candidate location by incorporating useful context information. Second, we propose to use an estimation network for predicting the cost from a candidate location to the destination. The two components are integrated in a principled way for deriving a more accurate cost of a candidate location for the A* algorithm. Extensive experiment results on three real-world datasets have shown the effectiveness and robustness of the proposed model.

Participant selection is a fundamental research issue in Mobile Crowdsensing (MCS). Previous approaches commonly assume that adequately long periods of candidate participants' historical mobility trajectories are available to model their patterns before the selection process, which is not realistic for some new MCS applications or platforms. The sparsity or even absence of mobility traces will incur inaccurate location prediction, thus undermining the deployment of new MCS applications. To this end, this paper investigates a novel problem called From-Scratch MCS (FS-MCS for short), in which we study how to intelligently select participants to minimize such cold-start effect. Specifically, we propose a novel framework based on reinforcement learning, named RL-Recruiter+. With the gradual accumulation of mobility trajectories over time, RL-Recruiter+ is able to make a good sequence of participant selection decisions for each sensing slot. Compared to its previous version, RL-Recruiter, Re-Recruiter+ jointly considers both the previous coverage and current mobility predictability when training the participant selection decision model. We evaluate our approach experimentally based on two real-world mobility datasets. The results demonstrate that RL-Recruiter+ outperforms the baseline approaches, including RL-Recruiter under various settings.

With the increasing accumulation of spatial-temporal trajectory data, location-based data mining has recently been extensively studied. A fundamental research topic in this field is learning the embedding vectors of locations through self-supervised pre-training. Pre-trained embedding vectors can utilize the highly available unlabeled trajectory data, and benefit downstream tasks in multiple aspects. However, most existing methods ignore the temporal information hidden in the visited time of locations in trajectories. Considering that human activities are highly regulated by specific periods of a day, temporal information can reflect some intrinsic characteristics of locations, so it is necessary to fuse them into location embedding vectors. In this paper, we propose a Time-Aware Location Embedding (TALE) pre-training method based on the CBOW framework, which is able to incorporate temporal information into the learned embedding vectors of locations. A novel temporal tree structure is designed to extract temporal information during the calculation of Hierarchical Softmax. In order to verify the effectiveness of TALE, we apply the learned embedding vectors into three downstream location-based prediction tasks, i.e., location classification, location visitor flow prediction and user next location prediction. Experiments are conducted on four real-world user trajectory datasets, and the experimental results demonstrate that our TALE model can obviously help downstream tasks gain better performance.

While massive efforts have been investigated in adversarial testing of convolutional neural networks (CNN), testing for recurrent neural networks (RNN) is still limited and leaves threats for vast sequential application domains. In this paper, we propose an adversarial testing framework RNN-Test for RNN systems, focusing on sequence-to-sequence (seq2seq) tasks of widespread deployments, not only classification domains. First, we design a novel search methodology customized for RNN models by maximizing the inconsistency of RNN states against their inner dependencies to produce adversarial inputs. Next, we introduce two state-based coverage metrics according to the distinctive structure of RNNs to exercise more system behaviors. Finally, RNN-Test solves the joint optimization problem to maximize state inconsistency and state coverage, and crafts adversarial inputs for various tasks of different kinds of inputs. For evaluations, we apply RNN-Test on four RNN models of common structures. On the tested models, the RNN-Test approach is demonstrated to be competitive in generating adversarial inputs, outperforming FGSM-based and DLFuzz-based methods to reduce the model performance more sharply with 2.78% to 37.94% higher success (or generation) rate. RNN-Test could also achieve 52.65% to 66.45% higher adversary rate than testRNN on MNIST LSTM model, as well as 53.76% to 58.02% more perplexity with 16% higher generation rate than DeepStellar on PTB language model. Compared with the traditional neuron coverage, the proposed state coverage metrics as guidance excel with 4.17% to 97.22% higher success (or generation) rate.

Crowdsourcing has become an effective tool to utilize human intelligence to perform tasks that are challenging for machines. Many truth discovery methods and incentive mechanisms for crowdsourcing have been proposed. However, most of them cannot deal with the crowdsourcing with copiers, who copy a part (or all) of data from other workers. This paper aims at designing crowdsourcing incentive mechanism for truth discovery of textual answers with copiers. We formulate the problem of maximizing the social welfare such that all tasks can be completed with the least confidence for truth discovery and design an three-stage incentive mechanism. In contextual embedding and clustering stage, we construct and cluster the content vector representations of textual crowdsourced answers at the semantic level. In truth discovery stage, we estimate the truth for each task based on the dependence and accuracy of workers. In reverse auction stage, we design a greedy algorithm to select the winners and determine the payment. Through both rigorous theoretical analysis and extensive simulations, we demonstrate that the proposed mechanisms achieve computational efficiency, individual rationality, truthfulness, and guaranteed approximation. Moreover, our truth discovery methods show prominent advantage in terms of precision when there are copiers in the crowdsourcing systems.

Event sequences are central to the analysis of data in domains that range from biology and health, to logfile analysis and people's everyday behavior. Many visualization tools have been created for such data, but people are error-prone when asked to judge the similarity of event sequences with basic presentation methods. This article describes an experiment that investigates whether local and global alignment techniques improve people's performance when judging sequence similarity. Participants were divided into three groups (basic versus local versus global alignment), and each participant judged the similarity of 180 sets of pseudo-randomly generated sequences. Each set comprised a target, a correct choice and a wrong choice. After training, the global alignment group was more accurate than the local alignment group (98 versus 93 percent correct), with the basic group getting 95 percent correct. Participants' response times were primarily affected by the number of event types, the similarity of sequences (measured by the Levenshtein distance) and the edit types (nine combinations of deletion, insertion and substitution). In summary, global alignment is superior and people's performance could be further improved by choosing alignment parameters that explicitly penalize sequence mismatches.

Few-shot learning under the N-way K-shot setting (i.e., K annotated samples for each of N classes) has been widely studied in relation extraction (e.g., FewRel) and image classification (e.g., Mini-ImageNet). Named entity recognition (NER) is typically framed as a sequence labeling problem where the entity classes are inherently entangled together because the entity number and classes in a sentence are not known in advance, leaving the N-way K-shot NER problem so far unexplored. In this paper, we first formally define a more suitable N-way K-shot setting for NER. Then we propose FewNER, a novel meta-learning approach for few-shot NER. FewNER separates the entire network into a task-independent part and a task-specific part. During training in FewNER, the task-independent part is meta-learned across multiple tasks and a task-specific part is learned for each single task in a low-dimensional space. At test time, FewNER keeps the task-independent part fixed and adapts to a new task via gradient descent by updating only the task-specific part, resulting in it being less prone to overfitting and more computationally efficient. The results demonstrate that FewNER achieves state-of-the-art performance against nine baseline methods by significant margins on three adaptation experiments.

In this article, fault diagnosis is of great significance for system health maintenance. For real applications, diagnosis accuracy suffers from unbalanced data patterns, where normal data are usually abundant than anomaly ones, leading to tremendous diagnosis obstacles. Therefore, it is challenging to use only normal data for fault diagnosis under this imbalanced condition. In addition, a single fault diagnosis model can only conduct one fault diagnosis task in most of cases. Accordingly, a one-class generative adversarial detection (OCGAD) framework based on semisupervised learning is proposed to learn one-class latent knowledge for dealing with multiple semisupervised fault diagnosis tasks, i.e., fault detection using only normal knowledge learning, novelty detection from unknown conditional data, and fault classification with unlabeled data. A bi-directional generative adversarial network (Bi-GAN) is first trained with only normal data. A one-class support vector machine is then established using features exacted by Bi-GAN from signals acquired from an attitude sensor for multifunctional fault detection. The presented OCGAD model is validated using an industrial robot with experiments of three fault detection tasks. The results demonstrate that the present model has good performance for dealing with multiple semisupervised diagnosis problems. info:eu-repo/semantics/publishedVersion

Workers participating in a crowdsourcing platform can have a wide range of abilities and interests. An important problem in crowdsourcing is the task recommendation problem, in which tasks that best match a particular worker's preferences and reliabilities are recommended to that worker. A task recommendation scheme that assigns tasks more likely to be accepted by a worker who is more likely to complete it reliably results in better performance for the task requester. Without prior information about a worker, his preferences and reliabilities need to be learned over time. In this paper, we propose a multi-armed bandit (MAB) framework to learn a worker's preferences and his reliabilities for different categories of tasks. However, unlike the classical MAB problem, the reward from the worker's completion of a task is unobservable. We therefore include the use of gold tasks (i.e., tasks whose solutions are known \emph{a priori} and which do not produce any rewards) in our task recommendation procedure. Our model could be viewed as a new variant of MAB, in which the random rewards can only be observed at those time steps where gold tasks are used, and the accuracy of estimating the expected reward of recommending a task to a worker depends on the number of gold tasks used. We show that the optimal regret is $O(\sqrt{n})$, where $n$ is the number of tasks recommended to the worker. We develop three task recommendation strategies to determine the number of gold tasks for different task categories, and show that they are order optimal. Simulations verify the efficiency of our approaches. Agency for Science, Technology and Research (A*STAR) Ministry of Education (MOE) Accepted version This work was supported in part by the Singapore Ministry of Education Academic Research Fund Tier 2 grant MOE2018-T2-2-019 and by A*STAR under its RIE2020 Advanced Manufacturing and Engineering (AME) Industry Alignment Fund – Pre Positioning (IAF-PP) (Grant No. A19D6a0053).

A content recommendation system based on human brain activity has become a reality. However, the cost of collecting the information from people is problematic. This article proposes a scheme that resolves the tradeoff between the inference performance from a profile model to a brain model and the cost of collecting profile information. In the proposed scheme, a machine learning model infers the brain model from the profile model and a feature selection method is applied to reduce the cost, i.e., the number of questionnaire items, of collecting profile information. Since only the top questionnaire items with the highest importance scores are used, we can maintain the inference performance as high as possible while limiting the number of questionnaire items. We demonstrate the effectiveness of the proposed scheme with a performance evaluation using an experimentally obtained brain model and a profile model created from real profile information. The results over different experimental parameters, video lengths, and feature selection methods demonstrate that the proposed scheme successfully identifies the top questionnaire items that contribute most significantly to the inference of brain models.

The predictive business process monitoring is a family of online approaches to predict the unfolding of running traces basedon the knowledge learned from historical event logs. In this paper, we address the task of predicting the next trace activity from the completed events in a running trace. This is an important business capability as counting on accurate predictions of the future activities may allow companies to guarantee the higher utilization by acting proactively in anticipation. We propose a novel predictive process approach that couples multi-view learning and deep learning, in order to gain predictive accuracy by accounting for the variety ofinformation possibly recorded in event logs. Experiments with various benchmark event logs prove the effectiveness of the proposed approach compared to several recent state-of-the-art methods.

Recent years have seen a rapidly growing number of online reviews of products. As a result, it is often not possible for customers to go through each review before making purchase decisions. One way to address this problem is to build a system for automatically addressing the helpfulness of reviews and present only those reviews that are determined to be helpful by the system to an end user. The vast majority of existing approaches to the task of review helpfulness prediction are based on hand-crafted features, thus making system performance heavily dependent on the quality of these features. In light of this weakness, we propose a new model of review helpfulness prediction using a combination of Convolutional Neural Network (CNN) and TransE wherein hand-crafted features can also be incorporated to improve the output. Specifically, CNN enables us to learn the semantic information from a review and TransE is used to capture the relationship between different entities mentioned in the review. Experiments on the Amazon product review datasets demonstrate that our approach significantly outperforms the state of the art.

Recent neural architectures in sequence labeling have yielded state-of-the-art performance on single domain data such as newswires. However, they still suffer from (i) requiring massive amounts of training data to avoid overfitting; (ii) huge performance degradation when there is a domain shift in the data distribution between training and testing. To make a sequence labeling system more broadly useful, it is crucial to reduce its training data requirements and transfer knowledge to other domains. In this paper, we investigate the problem of domain adaptation for sequence labeling under homogeneous and heterogeneous settings. We propose MetaSeq, a novel meta-learning approach for domain adaptation in sequence labeling. Specifically, MetaSeq incorporates meta-learning and adversarial training strategies to encourage robust, general and transferable representations for sequence labeling. The key advantage of MetaSeq is that it is capable of adapting to new unseen domains with a small amount of annotated data from those domains. We extensively evaluate MetaSeq on named entity recognition, part-of-speech tagging and slot filling under homogeneous and heterogeneous settings. The experimental results show that MetaSeq achieves state-of-the-art performance against eight baselines. Impressively, MetaSeq surpasses the in-domain performance using only 16.17% and 7% of target domain data on average for homogeneous settings, and 34.76%, 24%, 22.5% of target domain data on average for heterogeneous settings.

In this paper, we consider the problem of federated learning (FL) with training data that are not independent and identically distributed (non-IID) across the clients. To cope with data heterogeneity, Iterative Federated Clustering Algorithm (IFCA) has been proposed. IFCA partitions the clients into a number of clusters and lets the clients in the same cluster optimize a shared model. However, in IFCA, the clusters are non-overlapping, which leads to an inefficient utilization of the local information since the knowledge of a client is used by only one cluster during each round. To capture the complex nature of real-world data, soft clustering methods with overlapping clusters have been proposed that attain superior performance over the hard ones. Motivated by this, we propose a new algorithm named Federated Learning with Soft Clustering (FLSC) by combining the strengths of soft clustering and IFCA, where the clients are partitioned into overlapping clusters and the information of each participating client is used by multiple clusters simultaneously during each round. The experimental results show that FLSC achieves better learning performance on the classification tasks on the MNIST and Fashion-MNIST datasets, compared with the state-of-the-art baseline methods, i.e., the global model method and IFCA.

Recommender systems aim to accurately and actively provide users with potentially interesting items (products, information or services). Deep reinforcement learning has been successfully applied to recommender systems, but still heavily suffer from data sparsity and cold-start in real-world tasks. In this work, we propose an effective way to address such issues by leveraging the pervasive social networks among users in the estimation of action-values (Q). Specifically, we develop a Social Attentive Deep Q-network (SADQN) to approximate the optimal action-value function based on the preferences of both individual users and social neighbors, by successfully utilizing a social attention layer to model the influence between them. Further, we propose an enhanced variant of SADQN, termed SADQN++, to model the complicated and diverse trade-offs between personal preferences and social influence for all involved users, making the agent more powerful and flexible in learning the optimal policies. The experimental results on real-world datasets demonstrate that the proposed SADQNs remarkably outperform the state-of-the-art deep reinforcement learning agents, with reasonable computation cost.

Face aging has attracted widespread attention in recent years, but most studies are based on the same emotional situation. Is the same person’s aging in different emotional situations the same? To solve the above confusion, this paper proposes a novel face aging model DEF-Net, which consists of two parts: different emotional learnings (Emotion-Net) and face aging (Age-Net). Given a target emotion category, DEF-Net first assists the image from the original dataset to learn the emotion features through Emotion-Net and the generated dataset is used as the inputs of Age-Net. At the same time, multiple loss functions are used to ensure that the crucial information of the original image is not lost. Secondly, Age-Net, which has been pre-trained on the original dataset, began to adopt the generated dataset to learn the aging distribution under different emotions. Designed loss functions are utilized to ensure that the realistic target images generated by Age-Net do not lose the learned emotional characteristics. Finally, extensive experiments are used to verify the performance of DEF-Net. Compared with other state-of-the-art methods: (1) DEF-Net can learn different facial emotions across different datasets and generate corresponding realistic aging images; (2) the results achieved by our DEF-Net are demonstrated to be better than those by the model that performs face aging first and then learns different emotional characteristics.

A Relational-Sequential dataset (or RS-dataset for short) contains records comprised of a patient's values in demographic attributes and their sequence of diagnosis codes. The task of clustering an RS-dataset is helpful for analyses ranging from pattern mining to classification. However, existing methods are not appropriate to perform this task. Thus, we initiate a study of how an RS-dataset can be clustered effectively and efficiently. We formalize the task of clustering an RS-dataset as an optimization problem. At the heart of the problem is a distance measure we design to quantify the pairwise similarity between records of an RS-dataset. Our measure uses a tree structure that encodes hierarchical relationships between records, based on their demographics, as well as an edit-distance-like measure that captures both the sequentiality and the semantic similarity of diagnosis codes. We also develop an algorithm which first identifies k representative records (centers), for a given k, and then constructs k clusters, each containing one center and the records that are closer to the center compared to other centers. Experiments using two Electronic Health Record datasets demonstrate that our algorithm constructs compact and well-separated clusters, which preserve meaningful relationships between demographics and sequences of diagnosis codes, while being efficient and scalable.

Made available in DSpace on 2021-06-25T10:19:27Z (GMT). No. of bitstreams: 0 Previous issue date: 2020-01-01 Deep learning architectures have been widely fostered throughout the last years, being used in a wide range of applications, such as object recognition, image reconstruction, and signal processing. Nevertheless, such models suffer from a common problem known as overfitting, which limits the network from predicting unseen data effectively. Regularization approaches arise in an attempt to address such a shortcoming. Among them, one can refer to the well-known Dropout, which tackles the problem by randomly shutting down a set of neurons and their connections according to a certain probability. Therefore, this approach does not consider any additional knowledge to decide which units should be disconnected. In this paper, we propose an energy-based Dropout (E-Dropout) that makes conscious decisions whether a neuron should be dropped or not. Specifically, we design this regularization method by correlating neurons and the model’s energy as an importance level for further applying it to energy-based models, such as Restricted Boltzmann Machines (RBMs). The experimental results over several benchmark datasets revealed the proposed approach’s suitability compared to the traditional Dropout and the standard RBMs. São Paulo State University, Sao Paulo, SP 17033360 Brazil (e-mail: mateus.roder@unesp.br). São Paulo State University, Sao Paulo, SP 17033360 Brazil (e-mail: gustavo.rosa@unesp.br). ARMTEC Tecnologia em Robótica, Fortaleza, /CE 60150000 Brazil (e-mail: victor.albuquerque@ieee.org). São Paulo State University, Sao Paulo, SP 17033360 Brazil (e-mail: andre.rossi@unesp.br). São Paulo State University, Sao Paulo, SP 17033360 Brazil (e-mail: joao.papa@unesp.br).

Moving toward exascale computing, the size of data stored and accessed by applications is ever increasing. However, traditional disk-based storage has not seen improvements that keep up with the explosion of data volume or the speed of processors. Multiple levels of non-volatile storage devices are being added to handle bursty I/O, however, moving data across the storage hierarchy can take longer than the data generation or analysis. Asynchronous I/O can reduce the impact of I/O latency as it allows applications to schedule I/O early and to check their status later. I/O is thus overlapped with application communication or computation or both, effectively hiding some or all of the I/O latency. POSIX and MPI-I/O provide asynchronous read and write operations, but lack the support for non-data operations such as file open and close. Users also have to manually manage data dependencies and use low-level byte offsets, which requires significant effort and expertise to adopt. In this article, we present an asynchronous I/O framework that supports all types of I/O operations, manages data dependencies transparently and automatically, provides implicit and explicit modes for application flexibility, and error information retrieval. We implemented these techniques in HDF5. Our evaluation of several benchmarks and application workloads demonstrates it effectiveness on hiding the I/O cost from the application.

Just-in-time (JIT) bug prediction is an effective quality assurance activity that identifies whether a code commit will introduce bugs into the mobile app, aiming to provide prompt feedback to practitioners for priority review. Since collecting sufficient labeled bug data is not always feasible for some mobile apps, one possible approach is to leverage cross-app models. In this work, we propose a new cross-triplet deep feature embedding method, called CDFE, for cross-app JIT bug prediction task. The CDFE method incorporates a state-of-the-art cross-triplet loss function into a deep neural network to learn high-level feature representation for the cross-app data. This loss function adapts to the cross-app feature learning task and aims to learn a new feature space to shorten the distance of commit instances with the same label and enlarge the distance of commit instances with different labels. In addition, this loss function assigns higher weights to losses caused by cross-app instance pairs than that by intra-app instance pairs, aiming to narrow the discrepancy of cross-app bug data. We evaluate our CDFE method on a benchmark bug dataset from 19 mobile apps with two effort-aware indicators. The experimental results on 342 cross-app pairs show that our proposed CDFE method performs better than 14 baseline methods.

Analyzing short texts infers discriminative and coherent latent topics that is a critical and fundamental task since many real-world applications require semantic understanding of short texts. Traditional long text topic modeling algorithms (e.g., PLSA and LDA) based on word co-occurrences cannot solve this problem very well since only very limited word co-occurrence information is available in short texts. Therefore, short text topic modeling has already attracted much attention from the machine learning research community in recent years, which aims at overcoming the problem of sparseness in short texts. In this survey, we conduct a comprehensive review of various short text topic modeling techniques proposed in the literature. We present three categories of methods based on Dirichlet multinomial mixture, global word co-occurrences, and self-aggregation, with example of representative approaches in each category and analysis of their performance on various tasks. We develop the first comprehensive open-source library, called STTM, for use in Java that integrates all surveyed algorithms within a unified interface, benchmark datasets, to facilitate the expansion of new methods in this research field. Finally, we evaluate these state-of-the-art methods on many real-world datasets and compare their performance against one another and versus long text topic modeling algorithm., arXiv admin note: text overlap with arXiv:1808.02215 by other authors

Refactoring detection is crucial for a variety of applications and tasks: (i) empirical studies about code evolution, (ii) tools for library API migration, (iii) code reviews and change comprehension. However, recent research has questioned the accuracy of the state-of-the-art refactoring mining tools, which poses threats to the reliability of the detected refactorings. Moreover, the majority of refactoring mining tools depend on code similarity thresholds. Finding universal threshold values that can work well for all projects, regardless of their architectural style, application domain, and development practices is extremely challenging. Therefore, in a previous work [1], we introduced the first refactoring mining tool that does not require any code similarity thresholds to operate. In this work, we extend our tool to support low-level refactorings that take place within the body of methods. To evaluate our tool, we created one of the most accurate, complete, and representative refactoring oracles to date, including 7,223 true instances for 40 different refactoring types detected by one (minimum) up to six (maximum) different tools, and validated by one up to four refactoring experts. Our evaluation showed that our approach achieves the highest average precision (99.6%) and recall (94%) among all competitive tools, and on median is 2.6 times faster than the second faster competitive tool.

In this work, we formulate the problem of estimating and selecting task-relevant temporal signal segments from a single EEG trial in the form of a Markov decision process and propose a novel reinforcement-learning mechanism that can be combined with the existing deep-learning based BCI methods. To be specific, we devise an actor-critic network such that an agent can determine which timepoints need to be used (informative) or discarded (uninformative) in composing the intention-related features in a given trial, and thus enhancing the intention identification performance. To validate the effectiveness of our proposed method, we conducted experiments with a publicly available big MI dataset and applied our novel mechanism to various recent deep-learning architectures designed for MI classification. Based on the exhaustive experiments, we observed that our proposed method helped achieve statistically significant improvements in performance., 8 pages, 6 figures, 2 tables, and under review

The key to the effective control of a diffusion system lies in how accurately we could predict its unfolding dynamics based on the observation of its current state. However, in the real-world applications, it is often infeasible to conduct a timely and yet comprehensive observation due to resource constraints. In view of such a practical challenge, the goal of this work is to develop a novel computational method for performing active observations, termed active surveillance, with limited resources. Specifically, we aim to predict the dynamics of a large spatio-temporal diffusion system based on the observations of some of its components. Towards this end, we introduce a novel measure, the γ value, that enables us to identify the key components by means of modeling a sentinel network with a row sparsity structure. Having obtained a theoretical understanding of the γ value, we design a backward-selection sentinel network mining algorithm (SNMA) for deriving the sentinel network via group sparse Bayesian learning. In order to be practically useful, we further address the issue of scalability in the computation of SNMA, and moreover, extend SNMA to the case of a non-linear dynamical system that could involve complex diffusion mechanisms. We show the effectiveness of SNMA by validating it using both synthetic datasets and five real-world datasets. The experimental results are appealing, which demonstrate that SNMA readily outperforms the state-of-the-art methods.

Developing and optimizing software applications for high performance and energy efficiency is a very challenging task, even when considering a single target machine. For instance, optimizing for multicore-based computing systems requires in-depth knowledge about programming languages, application programming interfaces, compilers, performance tuning tools, and computer architecture and organization. Many of the tasks of performance engineering methodologies require manual efforts and the use of different tools not always part of an integrated toolchain. This paper presents Pegasus, a performance engineering approach supported by a framework that consists of a source-to-source compiler, controlled and guided by strategies programmed in a Domain-Specific Language, and an autotuner. Pegasus is a holistic and versatile approach spanning various decision layers composing the software stack, and exploiting the system capabilities and workloads effectively through the use of runtime autotuning. The Pegasus approach helps developers by automating tasks regarding the efficient implementation of software applications in multicore computing systems. These tasks focus on application analysis, profiling, code transformations, and the integration of runtime autotuning. Pegasus allows developers to program their strategies or to automatically apply existing strategies to software applications in order to ensure the compliance of non-functional requirements, such as performance and energy efficiency. We show how to apply Pegasus and demonstrate its applicability and effectiveness in a complex case study, which includes tasks from a smart navigation system.

Classifiers are among the most widely used supervised machine learning algorithms. Many classification models exist, and choosing the right one for a given task is difficult. During model selection and debugging, data scientists need to assess classifiers' performances, evaluate their learning behavior over time, and compare different models. Typically, this analysis is based on single-number performance measures such as accuracy. A more detailed evaluation of classifiers is possible by inspecting class errors. The confusion matrix is an established way for visualizing these class errors, but it was not designed with temporal or comparative analysis in mind. More generally, established performance analysis systems do not allow a combined temporal and comparative analysis of class-level information. To address this issue, we propose ConfusionFlow, an interactive, comparative visualization tool that combines the benefits of class confusion matrices with the visualization of performance characteristics over time. ConfusionFlow is model-agnostic and can be used to compare performances for different model types, model architectures, and/or training and test datasets. We demonstrate the usefulness of ConfusionFlow in a case study on instance selection strategies in active learning. We further assess the scalability of ConfusionFlow and present a use case in the context of neural network pruning., Changes compared to previous version: Reintroduced NN pruning use case; restructured Evaluation section; several additional minor revisions. Submitted as Minor Revision to IEEE TVCG on 2020-07-02

Deep learning (DL) based diagnosis models have to be trained by large quantities of monitoring data of machines. However, in real-case scenarios, machines operate under the normal condition in most of their life time while faults seldom happen. Therefore, though massive data are accessible, most are data of the normal condition while fault data are still extremely limited. In other words, fault diagnosis of real machines is actually a few-shot diagnosis problem. To deal with few-shot diagnosis, this article proposes adaptive knowledge transfer with multiclassifier ensemble (AKTME) under the paradigm of continual machine learning. In AKTME, knowledge learned by DL models is considered to be represented by the learnable filter kernels (FKs). The key of AKTME is a proposed continual weighted updating (CWU) technique of FKs. By CWU, shared FKs are distilled from multiple auxiliary tasks and adaptively transferred to the target task. Then by multiclassifier ensemble, AKTME is able to recognize faults with few fault data accessible. AKTME is applied on two few-shot diagnosis cases. Results verify that AKTME achieves higher diagnosis accuracies than recently proposed methods. Moreover, AKTME tends to improve the diagnosis accuracy as it prelearns on more auxiliary tasks continually.

With the modernization of machine learning techniques in healthcare, different innovations including support vector machine (SVM) have predominantly played a major role in classifying lung cancer, predicting coronavirus disease 2019, and other diseases. In particular, our algorithm focuses on integrated datasets as compared with other existing works. In this study, parallel-based SVM (P-SVM) and multiclass-based multiple submodels (MMSM-SVM) were used to analyze the optimal classification of lung diseases. This analysis aimed to find the optimal classification of lung diseases with id and stages, such as key-value pairs in MapReduce combined with P-SVM and MMSVM for binary and multiclasses, respectively. For nonlinear classification, kernel clustering-based SVM embedded with multiple submodels was developed. Both algorithms were developed using Apache spark environment, and data for the analysis were retrieved from microscope lab, UCI, Kaggle, and General Thoracic surgery database along with some electronic health records related to various lung diseases to increase the dataset size to 5 GB. Performance measures were conducted using a 5 GB dataset with five nodes. Dataset size was finally increased, and task analysis and CPU utilization were measured. © 2021 Tech Science Press. All rights reserved.

OBJECTIVE: The aim was to develop a reliable surgical quality assurance system for 2-stage esophagectomy. This development was conducted during the pilot phase of the multicenter ROMIO trial, collaborating with international experts.SUMMARY OF BACKGROUND DATA: There is evidence that the quality of surgical performance in randomized controlled trials influences clinical outcomes, quality of lymphadenectomy and loco-regional recurrence.METHODS: Standardization of 2-stage esophagectomy was based on structured observations, semi-structured interviews, hierarchical task analysis, and a Delphi consensus process. This standardization provided the structure for the operation manual and video and photographic assessment tools. Reliability was examined using generalizability theory.RESULTS: Hierarchical task analysis for 2-stage esophagectomy comprised fifty-four steps. Consensus (75%) agreement was reached on thirty-nine steps, whereas fifteen steps had a majority decision. An operation manual and record were created. A thirty five-item video assessment tool was developed that assessed the process (safety and efficiency) and quality of the end product (anatomy exposed and lymphadenectomy performed) of the operation. The quality of the end product section was used as a twenty seven-item photographic assessment tool. Thirty-one videos and fifty-three photographic series were submitted from the ROMIO pilot phase for assessment. The overall G-coefficient for the video assessment tool was 0.744, and for the photographic assessment tool was 0.700.CONCLUSIONS: A reliable surgical quality assurance system for 2-stage esophagectomy has been developed for surgical oncology randomized controlled trials.ETHICAL APPROVAL: 11/NW/0895 and confirmed locally as appropriate, 12/SW/0161, 16/SW/0098.TRIAL REGISTRATION NUMBER: ISRCTN59036820, ISRCTN10386621.

Assist-as-needed (AAN) control strategy is applied in robot-aided therapy to promote the subject's engagement with minimal assistance. Current methods are based on physical measurements or employ a model to estimate the subject's engagement, which indirectly reflects the physiology of the subject or needs a complicated modeling process. Besides, manual adjustments are required to tune the parameters of the AAN controller to adapt to the estimated engagement of subjects. Such two drawbacks of previous studies have limited their effectiveness for promoting the subject's engagement during the training. In this paper, we propose an engagement enhancement method based on Bayesian Optimization for the adaptive AAN controller. Firstly, we evaluate the subject's engagement by using sEMG-based muscle activation, which directly relates to the physiology of the subject and eliminates the need for complicated modeling of engagement. Next, Bayesian optimization, which is tolerant of noisy measurements and human adaptation, is used to search for the optimal parameter of the AAN controller for each training trial. Finally, we conduct a study with sixteen healthy human subjects to verify the effectiveness of the proposed method. The experimental results showed that subjects’ engagement in the experiment group improved more significantly and was maintained at a higher level than that in the control group during the training. The performance improvement after the training obtained in the experiment group was higher than that in the control group (45.42% v.s. 30.40%). The experimental results indicate that the proposed method may be promising for transferring to the rehabilitation of post-stroke patients.

Hard disk drive failures are one of the most common causes of service downtime in data centers. Predictive maintenance techniques have been adopted to extend the Remaining Useful Life (RUL) of these drives, and minimize service shortage and data loss. Several approaches based on machine and deep learning techniques have been proposed to address these issues, mostly exploiting models based on Self-Monitoring analysis and Reporting Technology (SMART) attributes. While these models have proven to be reliable, their performance is affected by the lack of information about the proximity of disk failure in time. Moreover, many of these techniques are sensitive to the highly unbalanced nature of existing data-sets, in terms of good to failed hard disk ratio. In this paper, we propose a LSTM based model combining SMART attributes and temporal analysis for estimating a hard drive health status according to its time to failure. Our approach outperforms state-of-the-art methods when evaluated on two data-sets, one containing hourly samples from 23,395 disks and the other reporting daily samples from 29,878 disks. Experimental results showed that our approach is well suited to data-sets with different sampling periods, being able to predict hard drive health status up to 45 days before failure.

In recent years, self-supervised paradigms for depth estimation have drawn lots of attention from the community. Promising as they are, in order to achieve wider application and better performance, reasoning about the uncertainty of the estimated depth maps is in great demand. In this paper, we propose a novel uncertainty-aware self-improving framework for self-supervised depth estimation, which could model and exploit uncertainty to boost depth accuracy without any external supervision. In detail, our framework includes two DepthNets–a teacher network and a student network. The teacher network is pre-trained in a self-supervised manner and the predicted depth is leveraged as pseudo-ground-truth to supervise the learning of the student network, leading to better estimation accuracy and the ability of modeling depth uncertainty. Afterward, the estimated uncertainty would be introduced back into the teacher network as the mask, preventing erroneous regions from contributing to network training and depth estimation. In this way, a better pseudo-ground-truth from the teacher network will be used for the next iteration of the student network training. We evaluate our proposed framework on the widely used KITTI dataset and TUM RGB-D dataset with both quantitative and qualitative experiments. The experimental results show that our framework achieves quite satisfactory performance among the self-supervised depth estimation methods. Extensive experiments which combine the student network and feature-based monocular SLAM system (i.e ORB-SLAM3) further demonstrate the benefits of the introduced depth uncertainty.

Multi-view semi-supervised learning (SSL) has attracted great attention due to its effectiveness in information utilization of multiple views and labeled and unlabeled data to solve practical problems. However, most existing methods exhibit high computational complexity. Effective integration of the information on different views to achieve enhanced performance remains a challenging task. In this study, we combine an anchor-based approach with multi-view semi-supervised learning to address these problems. A novel multi-view SSL method called fast multi-view SSL (FMSSL) based on learned graph is proposed. Starting from the affinity graphs constructed by using an anchor-based strategy, FMSSL learns an optimal multi-view consensus graph by using feature and label information. The learned graph can jointly consider the relation of multiple views to approximate the manifold structure. The learned graph is then introduced into the SSL model as the weight matrix of a bipartite graph to simultaneously perform separate classification on the original samples and anchors. Accordingly, multi-view SSL can be efficiently performed, and the computational complexity can be significantly reduced. We propose an effective algorithm to optimize the objective function. Extensive experimental results on different real-world datasets demonstrate the effectiveness and efficiency of the proposed algorithm.

Autonomous elevator operation is considered a promising solution for mobile navigation in office buildings. As a fundamental function, elevator button recognition remains unsolved due to the challenging image conditions and severe data imbalance problem. In this article, we propose an accurate and efficient framework, named OCR-RCNN, for elevator button recognition. The framework is comprised of an region-based convolutional neural network (R-CNN)-based button detector and an attention-RNN-based character recognizer. Leveraging the two components, we further propose an end-to-end architecture and a cascaded architecture to explore the most effective network design for the framework. Moreover, a perspective distortion removal algorithm is also developed to enhance the inference performance of OCR-RCNN. Another key contribution of this work is that we release the first large-scale elevator panel dataset with 2005 images and 21 767 button labels. Extensive experiments are conducted on the released dataset and other two publicly available datasets. The proposed framework achieves an F1 score of 0.94, 1.00, and 1.00 in detection task, and an accuracy of 79.6% 96.5%, and 96.4% in character recognition task. The results demonstrate the advantages of our method, outperforming alternative strategies and other state-of-the-art methods in the literature. The data and code are available on the project webpage https://github.com/zhudelong/ocr-rcnn-v2 .

The problem of continuous emotion recognition has been the subject of several studies. The proposed affective computing approaches employ sequential machine learning algorithms for improving the classification stage, accounting for the time ambiguity of emotional responses. Modeling and predicting the affective state over time is not a trivial problem because continuous data labeling is costly and not always feasible. This is a crucial issue in real-life applications, where data labeling is sparse and possibly captures only the most important events rather than the typical continuous subtle affective changes that occur. In this work, we introduce a framework from the machine learning literature called Multiple Instance Learning, which is able to model time intervals by capturing the presence or absence of relevant states, without the need to label the affective responses continuously (as required by standard sequential learning approaches). This choice offers a viable and natural solution for learning in a weakly supervised setting, taking into account the ambiguity of affective responses. We demonstrate the reliability of the proposed approach in a gold-standard scenario and towards real-world usage by employing an existing dataset (DEAP) and a purposely built one (Consumer). We also outline the advantages of this method with respect to standard supervised machine learning algorithms.

The study of human behaviors in cognitive sciences provides clues to understand and describe people’s personal and interpersonal functioning. In particular, the temporal analysis of behavioral dynamics can be a powerful tool to reveal events, correlations and causalities but also to discover abnormal behaviors. However, the annotation of these dynamics can be expensive in terms of temporal and human resources. To tackle this challenge, this paper proposes a methodology to semi-automatically annotate behavioral data. Behavioral dynamics can be expressed as sequences of simple dynamical processes: transitions between such processes are generally known as change-points. This paper describes the necessary steps to detect and classify change-points in behavioral data by using a dataset collected in a real use-case scenario. This dataset includes motor observations from children with typical development and with neuro-developmental disorders. Abnormal movements which are present in such disorders are useful to validate the system in conditions that are challenging even for experienced annotators. Results show that the system: can be effective in the semi-automated annotation task; can be efficient in presence of abnormal behaviors; may achieve good performance when trained with limited manually annotated data.