arXiv subjects

Di Chen

Publications and source records attributed to Di Chen.

69 records · Page 4Linked to original sources

arXiv2018

Learning Compact Recurrent Neural Networks with Block-Term Tensor Decomposition

Recurrent Neural Networks (RNNs) are powerful sequence modeling tools. However, when dealing with high dimensional inputs, the training of RNNs becomes computational expensive due to the large number of model parameters. This hinders RNNs from solving many important computer vision tasks, such as Action Recognition in Videos and Image Captioning. To overcome this problem, we propose a compact and flexible structure, namely Block-Term tensor decomposition, which greatly reduces the parameters of RNNs and improves their training efficiency. Compared with alternative low-rank approximations, such as tensor-train RNN (TT-RNN), our method, Block-Term RNN (BT-RNN), is not only more concise (when using the same rank), but also able to attain a better approximation to the original RNNs with much fewer parameters. On three challenging tasks, including Action Recognition in Videos, Image Captioning and Image Generation, BT-RNN outperforms TT-RNN and the standard RNN in terms of both prediction accuracy and convergence rate. Specifically, BT-LSTM utilizes 17,388 times fewer parameters than the standard LSTM to achieve an accuracy improvement over 15.6\% in the Action Recognition task on the UCF11 dataset.

cs.LG↗

arXiv2018

AV1 Video Coding Using Texture Analysis With Convolutional Neural Networks

Modern video codecs including the newly developed AOM/AV1 utilize hybrid coding techniques to remove spatial and temporal redundancy. However, efficient exploitation of statistical dependencies measured by a mean squared error (MSE) does not always produce the best psychovisual result. One interesting approach is to only encode visually relevant information and use a different coding method for "perceptually insignificant" regions in the frame, which can lead to substantial data rate reductions while maintaining visual quality. In this paper, we introduce a texture analyzer before encoding the input sequences to identify detail irrelevant texture regions in the frame using convolutional neural networks. We designed and developed a new coding tool referred to as texture mode for AV1, where if texture mode is selected at the encoder, no inter-frame prediction is performed for the identified texture regions. Instead, displacement of the entire region is modeled by just one set of motion parameters. Therefore, only the model parameters are transmitted to the decoder for reconstructing the texture regions. Non-texture regions in the frame are coded conventionally. We show that for many standard test sets, the proposed method achieved significant data rate reductions.

eess.IV↗

arXiv2018

Minmax Centered k-Partitioning of Trees and Applications to Sink Evacuation with Dynamic Confluent Flows

Let $T=(V,E)$ be a tree with associated costs on its subtrees. A minmax $k$-partition of $T$ is a partition into $k$ subtrees, minimizing the maximum cost of a subtree over all possible partitions. In the centered version of the problem, the cost of a subtree cost is defined as the minimum cost of "servicing" that subtree using a center located within it. The problem motivating this work was the sink-evacuation problem on trees, i.e., finding a collection of $k$-sinks that minimize the time required by a confluent dynamic network flow to evacuate all supplies to sinks. This paper provides the first polynomial-time algorithm for solving this problem, running in $O\Bigl(\max(k,\log n) k n \log^4 n\Bigr)$ time. The technique developed can be used to solve any Minmax Centered $k$-Partitioning problem on trees in which the servicing costs satisfy some very general conditions. Solutions can be found for both the discrete case, in which centers must be on vertices, and the continuous case, in which centers may also be placed on edges. The technique developed also improves previous results for finding a minmax cost $k$-partition of a tree given the location of the sinks in advance.

cs.DS↗

arXiv2018

Glassy Dynamics in a heavy ion irradiated NbSe2 crystal

Fascination with glassy states has persisted since Fisher introduced the vortex-glass as a new thermodynamic phase that is a true superconductor that lacks conventional long-range order. Though Fisher's original model considered point disorder, it was later predicted that columnar defects (CDs) could also induce glassiness -- specifically, a Bose-glass phase. In YBa$_2$Cu$_3$O$_{7-x}$ (YBCO), glassy states can cause distinct behavior in the temperature ($T$) dependent rate of thermally activated vortex motion ($S$). The vortex-glass state produces a plateau in $S(T)$ whereas a Bose-glass can transition into a state hosting vortex excitations called double-kinks that can expand, creating a large peak in $S(T)$. Although glass phases have been well-studied in YBCO, few studies exist of other materials containing CDs that could contribute to distinguishing universal behavior. Here, we report on the effectiveness of CDs tilted ~30$°$ from the c-axis in reducing $S$ in a NbSe$_2$ crystal. The magnetization is 5 times higher and $S$ is minimized when the field is parallel to the defects versus aligned with the c-axis. We see signatures of glassiness in both field orientations, but do not observe a peak in $S(T)$ nor a plateau at values observed in YBCO. We discuss the possibility that competing disorder induces a field-orientation-driven transition from a Bose-glass to an anisotropic glass involving both point and columnar disorder.

cond-mat.supr-con↗

arXiv2018

Building Better Quality Predictors Using "$ε$-Dominance"

Despite extensive research, many methods in software quality prediction still exhibit some degree of uncertainty in their results. Rather than treating this as a problem, this paper asks if this uncertainty is a resource that can simplify software quality prediction. For example, Deb's principle of $ε$-dominance states that if there exists some $ε$ value below which it is useless or impossible to distinguish results, then it is superfluous to explore anything less than $ε$. We say that for "large $ε$ problems", the results space of learning effectively contains just a few regions. If many learners are then applied to such large $ε$ problems, they would exhibit a "many roads lead to Rome" property; i.e., many different software quality prediction methods would generate a small set of very similar results. This paper explores DART, an algorithm especially selected to succeed for large $ε$ software quality prediction problems. DART is remarkable simple yet, on experimentation, it dramatically out-performs three sets of state-of-the-art defect prediction methods. The success of DART for defect prediction begs the questions: how many other domains in software quality predictors can also be radically simplified? This will be a fruitful direction for future work.

cs.SE↗

arXiv2018

Applications of Psychological Science for Actionable Analytics

Actionable analytics are those that humans can understand, and operationalize. What kind of data mining models generate such actionable analytics? According to psychological scientists, humans understand models that most match their own internal models, which they characterize as lists of "heuristic" (i.e., lists of very succinct rules). One such heuristic rule generator is the Fast-and-Frugal Trees (FFT) preferred by psychological scientists. Despite their successful use in many applied domains, FFTs have not been applied in software analytics. Accordingly, this paper assesses FFTs for software analytics. We find that FFTs are remarkably effective. Their models are very succinct (5 lines or less describing a binary decision tree). These succinct models outperform state-of-the-art defect prediction algorithms defined by Ghortra et al. at ICSE'15. Also, when we restrict training data to operational attributes (i.e., those attributes that are frequently changed by developers), FFTs perform much better than standard learners. Our conclusions are two-fold. Firstly, there is much that software analytics community could learn from psychological science. Secondly, proponents of complex methods should always baseline those methods against simpler alternatives. For example, FFTs could be used as a standard baseline learner against which other software analytics tools are compared.

cs.SE↗

arXiv2018

Multi-Reference Video Coding Using Stillness Detection

Encoders of AOM/AV1 codec consider an input video sequence as succession of frames grouped in Golden-Frame (GF) groups. The coding structure of a GF group is fixed with a given GF group size. In the current AOM/AV1 encoder, video frames are coded using a hierarchical, multilayer coding structure within one GF group. It has been observed that the use of multilayer coding structure may result in worse coding performance if the GF group presents consistent stillness across its frames. This paper proposes a new approach that adaptively designs the Golden-Frame (GF) group coding structure through the use of stillness detection. Our new approach hence develops an automatic stillness detection scheme using three metrics extracted from each GF group. It then differentiates those GF groups of stillness from other non- still GF groups and uses different GF coding structures accordingly. Experimental result demonstrates a consistent coding gain using the new approach.

eess.IV↗

arXiv2018

Texture Segmentation Based Video Compression Using Convolutional Neural Networks

There has been a growing interest in using different approaches to improve the coding efficiency of modern video codec in recent years as demand for web-based video consumption increases. In this paper, we propose a model-based approach that uses texture analysis/synthesis to reconstruct blocks in texture regions of a video to achieve potential coding gains using the AV1 codec developed by the Alliance for Open Media (AOM). The proposed method uses convolutional neural networks to extract texture regions in a frame, which are then reconstructed using a global motion model. Our preliminary results show an increase in coding efficiency while maintaining satisfactory visual quality.

cs.CV↗

arXiv2017

BT-Nets: Simplifying Deep Neural Networks via Block Term Decomposition

Recently, deep neural networks (DNNs) have been regarded as the state-of-the-art classification methods in a wide range of applications, especially in image classification. Despite the success, the huge number of parameters blocks its deployment to situations with light computing resources. Researchers resort to the redundancy in the weights of DNNs and attempt to find how fewer parameters can be chosen while preserving the accuracy at the same time. Although several promising results have been shown along this research line, most existing methods either fail to significantly compress a well-trained deep network or require a heavy fine-tuning process for the compressed network to regain the original performance. In this paper, we propose the \textit{Block Term} networks (BT-nets) in which the commonly used fully-connected layers (FC-layers) are replaced with block term layers (BT-layers). In BT-layers, the inputs and the outputs are reshaped into two low-dimensional high-order tensors, then block-term decomposition is applied as tensor operators to connect them. We conduct extensive experiments on benchmark datasets to demonstrate that BT-layers can achieve a very large compression ratio on the number of parameters while preserving the representation power of the original FC-layers as much as possible. Specifically, we can get a higher performance while requiring fewer parameters compared with the tensor train method.

stat.ML↗

arXiv2017

Channel Matrix Sparsity with Imperfect Channel State Information in Cloud-Radio Access Networks

Channel matrix sparsification is considered as a promising approach to reduce the progressing complexity in large-scale cloud-radio access networks (C-RANs) based on ideal channel condition assumption. In this paper, the research of channel sparsification is extend to practical scenarios, in which the perfect channel state information (CSI) is not available. First, a tractable lower bound of signal-to-interferenceplus-noise ratio (SINR) fidelity, which is defined as a ratio of SINRs with and without channel sparsification, is derived to evaluate the impact of channel estimation error. Based on the theoretical results, a Dinkelbach-based algorithm is proposed to achieve the global optimal performance of channel matrix sparsification based on the criterion of distance. Finally, all these results are extended to a more challenging scenario with pilot contamination. Finally, simulation results are shown to evaluate the performance of channel matrix sparsification with imperfect CSIs and verify our analytical results.

cs.IT↗

arXiv2017

Multi-Entity Dependence Learning with Rich Context via Conditional Variational Auto-encoder

Multi-Entity Dependence Learning (MEDL) explores conditional correlations among multiple entities. The availability of rich contextual information requires a nimble learning scheme that tightly integrates with deep neural networks and has the ability to capture correlation structures among exponentially many outcomes. We propose MEDL_CVAE, which encodes a conditional multivariate distribution as a generating process. As a result, the variational lower bound of the joint likelihood can be optimized via a conditional variational auto-encoder and trained end-to-end on GPUs. Our MEDL_CVAE was motivated by two real-world applications in computational sustainability: one studies the spatial correlation among multiple bird species using the eBird data and the other models multi-dimensional landscape composition and human footprint in the Amazon rainforest with satellite images. We show that MEDL_CVAE captures rich dependency structures, scales better than previous methods, and further improves on the joint likelihood taking advantage of very large datasets that are beyond the capacity of previous methods.

cs.LG↗

arXiv2017

Replicating and Scaling up Qualitative Analysis using Crowdsourcing: A Github-based Case Study

Due to the difficulties in replicating and scaling up qualitative studies, such studies are rarely verified. Accordingly, in this paper, we leverage the advantages of crowdsourcing (low costs, fast speed, scalable workforce) to replicate and scale-up one state-of-the-art qualitative study. That qualitative study explored 20 GitHub pull requests to learn factors that influence the fate of pull requests with respect to approval and merging. As a secondary study, using crowdsourcing at a cost of $200, we studied 250 pull requests from 142 GitHub projects. The prior qualitative findings are mapped into questions for crowds workers. Their answers were converted into binary features to build a predictor which predicts whether code would be merged with median F1 scores of 68%. For the same large group of pull requests, the median F1 scores could achieve 90% by a predictor built with additional features defined by prior quantitative results. Based on this case study, we conclude that there is much benefit in combining different kinds of research methods. While qualitative insights are very useful for finding novel insights, they can be hard to scale or replicate. That said, they can guide and define the goals of scalable secondary studies that use (e.g.) crowdsourcing+data mining. On the other hand, while data mining methods are reproducible and scalable to large data sets, their results may be spectacularly wrong since they lack contextual information. That said, they can be used to test the stability and external validity, of the insights gained from a qualitative analysis.

cs.SE↗

arXiv2017

Deep Multi-Species Embedding

Understanding how species are distributed across landscapes over time is a fundamental question in biodiversity research. Unfortunately, most species distribution models only target a single species at a time, despite strong ecological evidence that species are not independently distributed. We propose Deep Multi-Species Embedding (DMSE), which jointly embeds vectors corresponding to multiple species as well as vectors representing environmental covariates into a common high-dimensional feature space via a deep neural network. Applied to bird observational data from the citizen science project \textit{eBird}, we demonstrate how the DMSE model discovers inter-species relationships to outperform single-species distribution models (random forests and SVMs) as well as competing multi-label models. Additionally, we demonstrate the benefit of using a deep neural network to extract features within the embedding and show how they improve the predictive performance of species distribution modelling. An important domain contribution of the DMSE model is the ability to discover and describe species interactions while simultaneously learning the shared habitat preferences among species. As an additional contribution, we provide a graphical embedding of hundreds of bird species in the Northeast US.

cs.LG↗

arXiv2016

Minmax Tree Facility Location and Sink Evacuation with Dynamic Confluent Flows

Let $G=(V,E)$ be a graph modelling a building or road network in which edges have-both travel times (lengths) and capacities associated with them. An edge's capacity is the number of people that can enter that edge in a unit of time. In emergencies, people evacuate towards the exits. If too many people try to evacuate through the same edge, congestion builds up and slows down the evacuation. Graphs with both lengths and capacities are known as Dynamic Flow networks. An evacuation plan for $G$ consists of a choice of exit locations and a partition of the people at the vertices into groups, with each group evacuating to the same exit. The evacuation time of a plan is the time it takes until the last person evacuates. The $k$-sink evacuation problem is to provide an evacuation plan with $k$ exit locations that minimizes the evacuation time. It is known that this problem is NP-Hard for general graphs but no polynomial time algorithm was previously known even for the case of $G$ a tree. This paper presents an $O(n k^2 \log^5 n)$ algorithm for the $k$-sink evacuation problem on trees. Our algorithms also apply to a more general class of problems, which we call minmax tree facility location.

cs.DS↗

arXiv2016

Backhaul Traffic Balancing and Dynamic Content-Centric Clustering for the Downlink of Fog Radio Access Network

Recently, an evolution of the Cloud Radio Access Network (C-RAN) has been proposed, named as Fog Radio Access Network (F-RAN). Compared to C-RAN, the Radio Units (RUs) in F-CAN are equipped with local caches, which can store some frequently requested files. In the downlink, users requesting the same file form a multicast group, and are cooperatively served by a cluster of RUs. The requested file is either available locally in the cache of this cluster or fetched from the Central Processor (CP) via backhauls. Thus caching some frequently requested files can greatly reduce the burden on backhaul links. Whether a specific RU should be involved in a cluster to serve a multicast group depends on its backhaul capacity, requested files, cached files and the channel. Therefore it is subject to optimization. In this paper we investigate the joint design of multicast beamforming, dynamic clustering and backhaul traffic balancing. Beamforming and clustering are jointly optimized in order to minimize the power consumed, while QoS of each user is to be met and the traffic on each backhaul link is balanced according to its capacity.

cs.IT↗