January 27, 2020

3044 words 15 mins read

Paper Group ANR 1205

Consistent Cross-view Matching for Unsupervised Person Re-identification. Bayesian Optimization for Materials Design with Mixed Quantitative and Qualitative Variables. Dynamic Network Embeddings for Network Evolution Analysis. Minimax bounds for structured prediction. Robust Anomaly Detection and Backdoor Attack Detection Via Differential Privacy. …

Consistent Cross-view Matching for Unsupervised Person Re-identification


Title	Consistent Cross-view Matching for Unsupervised Person Re-identification
Authors	Xueping Wang, Rameswar Panda, Min Liu, Amit K Roy-Chowdhury
Abstract	Most existing unsupervised person re-identificationmethods focus on learning an identity discriminative feature em-bedding for efficiently representing images of different persons.However, higher-order relationships across the entire cameranetwork are often ignored leading to contradictory outputs whenthe results of different camera pairs are combined. In this paper,we address this problem by proposing a consistent cross-viewmatching framework for unsupervised person re-identificationby exploiting more reliable positive image pairs in a cameranetwork. Specifically, we first construct a bipartite graph foreach pair of cameras, in which each node denotes a person, andthen graph matching is used to obtain optimal global matchesacross camera pairs. Thereafter, loop consistent and transitiveinference consistent constraints are introduced into the cross-view matches, which consider similarity relationshipsacross theentire camera networkto increase confidence in the matched/non-matched pairs. We then train distance metric models for eachcamera pair using the reliably matched image pairs. Finally,we embed the cross-view matching method into an iterativeupdating framework that iterates between the consistent cross-view matching and the cross-view distance metric learning. Wedemonstrate the superiority of the proposed method over thestate-of-the-art unsupervised person re-identification methodson three benchmark datasets such as Market1501, MARS andDukeMTMC-VideoReID datasets
Tasks	Graph Matching, Metric Learning, Person Re-Identification, Unsupervised Person Re-Identification
Published	2019-08-27
URL	https://arxiv.org/abs/1908.10486v1
PDF	https://arxiv.org/pdf/1908.10486v1.pdf
PWC	https://paperswithcode.com/paper/consistent-cross-view-matching-for
Repo
Framework

Bayesian Optimization for Materials Design with Mixed Quantitative and Qualitative Variables


Title	Bayesian Optimization for Materials Design with Mixed Quantitative and Qualitative Variables
Authors	Yichi Zhang, Daniel Apley, Wei Chen
Abstract	Although Bayesian Optimization (BO) has been employed for accelerating materials design in computational materials engineering, existing works are restricted to problems with quantitative variables. However, real designs of materials systems involve both qualitative and quantitative design variables representing material compositions, microstructure morphology, and processing conditions. For mixed-variable problems, existing Bayesian Optimization (BO) approaches represent qualitative factors by dummy variables first and then fit a standard Gaussian process (GP) model with numerical variables as the surrogate model. This approach is restrictive theoretically and fails to capture complex correlations between qualitative levels. We present in this paper the integration of a novel latent-variable (LV) approach for mixed-variable GP modeling with the BO framework for materials design. LVGP is a fundamentally different approach that maps qualitative design variables to underlying numerical LV in GP, which has strong physical justification. It provides flexible parameterization and representation of qualitative factors and shows superior modeling accuracy compared to the existing methods. We demonstrate our approach through testing with numerical examples and materials design examples. It is found that in all test examples the mapped LVs provide intuitive visualization and substantial insight into the nature and effects of the qualitative factors. Though materials designs are used as examples, the method presented is generic and can be utilized for other mixed variable design optimization problems that involve expensive physics-based simulations.
Tasks
Published	2019-10-03
URL	https://arxiv.org/abs/1910.01688v1
PDF	https://arxiv.org/pdf/1910.01688v1.pdf
PWC	https://paperswithcode.com/paper/bayesian-optimization-for-materials-design-1
Repo
Framework

Dynamic Network Embeddings for Network Evolution Analysis


Title	Dynamic Network Embeddings for Network Evolution Analysis
Authors	Chuanchang Chen, Yubo Tao, Hai Lin
Abstract	Network embeddings learn to represent nodes as low-dimensional vectors to preserve the proximity between nodes and communities of the network for network analysis. The temporal edges (e.g., relationships, contacts, and emails) in dynamic networks are important for network evolution analysis, but few existing methods in network embeddings can capture the dynamic information from temporal edges. In this paper, we propose a novel dynamic network embedding method to analyze evolution patterns of dynamic networks effectively. Our method uses random walk to keep the proximity between nodes and applies dynamic Bernoulli embeddings to train discrete-time network embeddings in the same vector space without alignments to preserve the temporal continuity of stable nodes. We compare our method with several state-of-the-art methods by link prediction and evolving node detection, and the experiments demonstrate that our method generally has better performance in these tasks. Our method is further verified by two real-world dynamic networks via detecting evolving nodes and visualizing their temporal trajectories in the embedded space.
Tasks	Link Prediction, Network Embedding
Published	2019-06-24
URL	https://arxiv.org/abs/1906.09860v1
PDF	https://arxiv.org/pdf/1906.09860v1.pdf
PWC	https://paperswithcode.com/paper/dynamic-network-embeddings-for-network
Repo
Framework

Minimax bounds for structured prediction


Title	Minimax bounds for structured prediction
Authors	Kevin Bello, Asish Ghoshal, Jean Honorio
Abstract	Structured prediction can be considered as a generalization of many standard supervised learning tasks, and is usually thought as a simultaneous prediction of multiple labels. One standard approach is to maximize a score function on the space of labels, which decomposes as a sum of unary and pairwise potentials, each depending on one or two specific labels, respectively. For this approach, several learning and inference algorithms have been proposed over the years, ranging from exact to approximate methods while balancing the computational complexity. However, in contrast to binary and multiclass classification, results on the necessary number of samples for achieving learning is still limited, even for a specific family of predictors such as factor graphs. In this work, we provide minimax bounds for a class of factor-graph inference models for structured prediction. That is, we characterize the necessary sample complexity for any conceivable algorithm to achieve learning of factor-graph predictors.
Tasks	Structured Prediction
Published	2019-06-02
URL	https://arxiv.org/abs/1906.00449v1
PDF	https://arxiv.org/pdf/1906.00449v1.pdf
PWC	https://paperswithcode.com/paper/190600449
Repo
Framework

Robust Anomaly Detection and Backdoor Attack Detection Via Differential Privacy


Title	Robust Anomaly Detection and Backdoor Attack Detection Via Differential Privacy
Authors	Min Du, Ruoxi Jia, Dawn Song
Abstract	Outlier detection and novelty detection are two important topics for anomaly detection. Suppose the majority of a dataset are drawn from a certain distribution, outlier detection and novelty detection both aim to detect data samples that do not fit the distribution. Outliers refer to data samples within this dataset, while novelties refer to new samples. In the meantime, backdoor poisoning attacks for machine learning models are achieved through injecting poisoning samples into the training dataset, which could be regarded as “outliers” that are intentionally added by attackers. Differential privacy has been proposed to avoid leaking any individual’s information, when aggregated analysis is performed on a given dataset. It is typically achieved by adding random noise, either directly to the input dataset, or to intermediate results of the aggregation mechanism. In this paper, we demonstrate that applying differential privacy can improve the utility of outlier detection and novelty detection, with an extension to detect poisoning samples in backdoor attacks. We first present a theoretical analysis on how differential privacy helps with the detection, and then conduct extensive experiments to validate the effectiveness of differential privacy in improving outlier detection, novelty detection, and backdoor attack detection.
Tasks	Anomaly Detection, Outlier Detection
Published	2019-11-16
URL	https://arxiv.org/abs/1911.07116v1
PDF	https://arxiv.org/pdf/1911.07116v1.pdf
PWC	https://paperswithcode.com/paper/robust-anomaly-detection-and-backdoor-attack-1
Repo
Framework

Node classification framework


Title	Node classification framework
Authors	Keting Cen, Huawei Shen, Jinhua Gao, Qi Cao, Bingbing Xu, Xueqi Cheng
Abstract	GCN is a recent effective algorithm which effectively learns a function incorporate both graph structure and node features for semisupervised graph based node classification. Although GCN exceeds other state-of-the-art methods, the number of parameters that need to be learned are still significantly more than the number of used samples. Because only labeled nodes and their 2-hop neighbors are used for learning parameters in GCN. As a consequence the accuracy of node classification declines sharply when decreasing the size of train data. In order to reducing parameters and making better use of unlabeled nodes, we proposes to use unsupervised representation algorithm such as autoencoder to pretrain a embedding for each node, which captures effective information from input features and reduce parameters in graph convolution layers. Experiments show that our model has increased by 0.87%, 1.96%, 1.87% compared to baseline on Cora, Citeseer and Pubmed data respectively. The performance of our model declines slower than GCN when decreasing the train samples.
Tasks	Link Prediction, Network Embedding, Node Classification
Published	2019-06-20
URL	https://arxiv.org/abs/1906.08745v2
PDF	https://arxiv.org/pdf/1906.08745v2.pdf
PWC	https://paperswithcode.com/paper/a-graph-auto-encoder-for-attributed-network
Repo
Framework

Resilient Combination of Complementary CNN and RNN Features for Text Classification through Attention and Ensembling


Title	Resilient Combination of Complementary CNN and RNN Features for Text Classification through Attention and Ensembling
Authors	Athanasios Giannakopoulos, Maxime Coriou, Andreea Hossmann, Michael Baeriswyl, Claudiu Musat
Abstract	State-of-the-art methods for text classification include several distinct steps of pre-processing, feature extraction and post-processing. In this work, we focus on end-to-end neural architectures and show that the best performance in text classification is obtained by combining information from different neural modules. Concretely, we combine convolution, recurrent and attention modules with ensemble methods and show that they are complementary. We introduce ECGA, an end-to-end go-to architecture for novel text classification tasks. We prove that it is efficient and robust, as it attains or surpasses the state-of-the-art on varied datasets, including both low and high data regimes.
Tasks	Text Classification
Published	2019-03-28
URL	http://arxiv.org/abs/1903.12157v1
PDF	http://arxiv.org/pdf/1903.12157v1.pdf
PWC	https://paperswithcode.com/paper/resilient-combination-of-complementary-cnn
Repo
Framework

A Comparison of Neural Network Training Methods for Text Classification


Title	A Comparison of Neural Network Training Methods for Text Classification
Authors	Anderson de Andrade
Abstract	We study the impact of neural networks in text classification. Our focus is on training deep neural networks with proper weight initialization and greedy layer-wise pretraining. Results are compared with 1-layer neural networks and Support Vector Machines. We work with a dataset of labeled messages from the Twitter microblogging service and aim to predict weather conditions. A feature extraction procedure specific for the task is proposed, which applies dimensionality reduction using Latent Semantic Analysis. Our results show that neural networks outperform Support Vector Machines with Gaussian kernels, noticing performance gains from introducing additional hidden layers with nonlinearities. The impact of using Nesterov’s Accelerated Gradient in backpropagation is also studied. We conclude that deep neural networks are a reasonable approach for text classification and propose further ideas to improve performance.
Tasks	Dimensionality Reduction, Text Classification
Published	2019-10-28
URL	https://arxiv.org/abs/1910.12674v1
PDF	https://arxiv.org/pdf/1910.12674v1.pdf
PWC	https://paperswithcode.com/paper/a-comparison-of-neural-network-training
Repo
Framework

A Comparison of Information Retrieval Techniques for Detecting Source Code Plagiarism


Title	A Comparison of Information Retrieval Techniques for Detecting Source Code Plagiarism
Authors	Vasishtha Sriram Jayapati, Ajay Venkitaraman
Abstract	Plagiarism is a commonly encountered problem in the academia. While there are several tools and techniques to efficiently determine plagiarism in text, the same cannot be said about source code plagiarism. To make the existing systems more efficient, we use several information retrieval techniques to find the similarity between source code files written in Java. We later use JPlag, which is a string-based plagiarism detection tool used in academia to match the plagiarized source codes. In this paper, we aim to generalize on the efficiency and effectiveness of detecting plagiarism using different information retrieval models rather than using just string manipulation algorithms.
Tasks	Information Retrieval
Published	2019-02-06
URL	http://arxiv.org/abs/1902.02407v1
PDF	http://arxiv.org/pdf/1902.02407v1.pdf
PWC	https://paperswithcode.com/paper/a-comparison-of-information-retrieval
Repo
Framework

Information theoretic learning of robust deep representations


Title	Information theoretic learning of robust deep representations
Authors	Nicolas Pinchaud
Abstract	We propose a novel objective function for learning robust deep representations of data based on information theory. Data is projected into a feature-vector space such that the mutual information of all subsets of features relative to the supervising signal is maximized. This objective function gives rise to robust representations by conserving available information relative to supervision in the face of noisy or unavailable features. Although the objective function is not directly tractable, we are able to derive a surrogate objective function. Minimizing this surrogate loss encourages features to be non-redundant and conditionally independent relative to the supervising signal. To evaluate the quality of obtained solutions, we have performed a set of preliminary experiments that show promising results.
Tasks
Published	2019-05-30
URL	https://arxiv.org/abs/1905.12874v1
PDF	https://arxiv.org/pdf/1905.12874v1.pdf
PWC	https://paperswithcode.com/paper/information-theoretic-learning-of-robust-deep
Repo
Framework

Prediction is very hard, especially about conversion. Predicting user purchases from clickstream data in fashion e-commerce


Title	Prediction is very hard, especially about conversion. Predicting user purchases from clickstream data in fashion e-commerce
Authors	Luca Bigon, Giovanni Cassani, Ciro Greco, Lucas Lacasa, Mattia Pavoni, Andrea Polonioli, Jacopo Tagliabue
Abstract	Knowing if a user is a buyer vs window shopper solely based on clickstream data is of crucial importance for ecommerce platforms seeking to implement real-time accurate NBA (next best action) policies. However, due to the low frequency of conversion events and the noisiness of browsing data, classifying user sessions is very challenging. In this paper, we address the clickstream classification problem in the fashion industry and present three major contributions to the burgeoning field of AI in fashion: first, we collected, normalized and prepared a novel dataset of live shopping sessions from a major European e-commerce fashion website; second, we use the dataset to test in a controlled environment strong baselines and SOTA models from the literature; finally, we propose a new discriminative neural model that outperforms neural architectures recently proposed at Rakuten labs.
Tasks
Published	2019-06-30
URL	https://arxiv.org/abs/1907.00400v1
PDF	https://arxiv.org/pdf/1907.00400v1.pdf
PWC	https://paperswithcode.com/paper/prediction-is-very-hard-especially-about
Repo
Framework

Masked Non-Autoregressive Image Captioning


Title	Masked Non-Autoregressive Image Captioning
Authors	Junlong Gao, Xi Meng, Shiqi Wang, Xia Li, Shanshe Wang, Siwei Ma, Wen Gao
Abstract	Existing captioning models often adopt the encoder-decoder architecture, where the decoder uses autoregressive decoding to generate captions, such that each token is generated sequentially given the preceding generated tokens. However, autoregressive decoding results in issues such as sequential error accumulation, slow generation, improper semantics and lack of diversity. Non-autoregressive decoding has been proposed to tackle slow generation for neural machine translation but suffers from multimodality problem due to the indirect modeling of the target distribution. In this paper, we propose masked non-autoregressive decoding to tackle the issues of both autoregressive decoding and non-autoregressive decoding. In masked non-autoregressive decoding, we mask several kinds of ratios of the input sequences during training, and generate captions parallelly in several stages from a totally masked sequence to a totally non-masked sequence in a compositional manner during inference. Experimentally our proposed model can preserve semantic content more effectively and can generate more diverse captions.
Tasks	Image Captioning, Machine Translation
Published	2019-06-03
URL	https://arxiv.org/abs/1906.00717v1
PDF	https://arxiv.org/pdf/1906.00717v1.pdf
PWC	https://paperswithcode.com/paper/190600717
Repo
Framework

DISCO: Influence Maximization Meets Network Embedding and Deep Learning


Title	DISCO: Influence Maximization Meets Network Embedding and Deep Learning
Authors	Hui Li, Mengting Xu, Sourav S Bhowmick, Changsheng Sun, Zhongyuan Jiang, Jiangtao Cui
Abstract	Since its introduction in 2003, the influence maximization (IM) problem has drawn significant research attention in the literature. The aim of IM is to select a set of k users who can influence the most individuals in the social network. The problem is proven to be NP-hard. A large number of approximate algorithms have been proposed to address this problem. The state-of-the-art algorithms estimate the expected influence of nodes based on sampled diffusion paths. As the number of required samples have been recently proven to be lower bounded by a particular threshold that presets tradeoff between the accuracy and efficiency, the result quality of these traditional solutions is hard to be further improved without sacrificing efficiency. In this paper, we present an orthogonal and novel paradigm to address the IM problem by leveraging deep learning models to estimate the expected influence. Specifically, we present a novel framework called DISCO that incorporates network embedding and deep reinforcement learning techniques to address this problem. Experimental study on real-world networks demonstrates that DISCO achieves the best performance w.r.t efficiency and influence spread quality compared to state-of-the-art classical solutions. Besides, we also show that the learning model exhibits good generality.
Tasks	Network Embedding
Published	2019-06-18
URL	https://arxiv.org/abs/1906.07378v1
PDF	https://arxiv.org/pdf/1906.07378v1.pdf
PWC	https://paperswithcode.com/paper/disco-influence-maximization-meets-network
Repo
Framework

Functional Representation of Large-Scale Heterogeneous RNA Sequences with Integration of Diverse Multi-omics, Interactions, and Annotations Data


Title	Functional Representation of Large-Scale Heterogeneous RNA Sequences with Integration of Diverse Multi-omics, Interactions, and Annotations Data
Authors	Nhat Tran, Jean Gao
Abstract	Long non-coding RNA, microRNA, and messenger RNA enable key regulations of various biological processes through a variety of diverse interaction mechanisms. Identifying the interactions and cross-talk between these heterogeneous RNA classes is essential in order to uncover the functional role of individual RNA transcripts, especially for unannotated and newly-discovered RNA sequences with no known interactions. Recently, sequence-based deep learning and network embedding methods are becoming promising approaches that can either predict RNA-RNA interactions from a sequence or infer missing interactions from patterns that may exist in the network topology. However, the majority of these methods have several limitations, eg, the inability to perform inductive predictions, to distinguish the directionality of interactions, or to integrate various sequence, interaction, and annotation biological datasets. We proposed a novel deep learning-based framework, rna2rna, which learns from RNA sequences to produce a low-dimensional embedding that preserves the proximities in both the interactions topology and the functional affinity topology. In this proposed embedding space, we have designated a two-part” source and target contexts” to capture the targeting and receptive fields of each RNA transcript, while encapsulating the heterogenous cross-talk interactions between lncRNAs and miRNAs. From experimental results, our method exhibits superior performance in AUPR rates compared to state-of-art approaches at predicting missing interactions in different RNA-RNA interaction databases and was shown to accurately perform link predictions to novel RNA sequences not seen at training time, even without any prior information. Additional results suggest that our proposed framework can capture a manifold for heterogeneous RNA sequences to discover novel functional annotations.
Tasks	Network Embedding
Published	2019-06-17
URL	https://arxiv.org/abs/1906.07289v2
PDF	https://arxiv.org/pdf/1906.07289v2.pdf
PWC	https://paperswithcode.com/paper/rna2rna-predicting-lncrna-microrna-mrna
Repo
Framework

The OoO VLIW JIT Compiler for GPU Inference


Title	The OoO VLIW JIT Compiler for GPU Inference
Authors	Paras Jain, Xiangxi Mo, Ajay Jain, Alexey Tumanov, Joseph E. Gonzalez, Ion Stoica
Abstract	Current trends in Machine Learning~(ML) inference on hardware accelerated devices (e.g., GPUs, TPUs) point to alarmingly low utilization. As ML inference is increasingly time-bounded by tight latency SLOs, increasing data parallelism is not an option. The need for better efficiency motivates GPU multiplexing. Furthermore, existing GPU programming abstractions force programmers to micro-manage GPU resources in an early-binding, context-free fashion. We propose a VLIW-inspired Out-of-Order (OoO) Just-in-Time (JIT) compiler that coalesces and reorders execution kernels at runtime for throughput-optimal device utilization while satisfying latency SLOs. We quantify the inefficiencies of space-only and time-only multiplexing alternatives and demonstrate an achievable 7.7x opportunity gap through spatial coalescing.
Tasks
Published	2019-01-28
URL	http://arxiv.org/abs/1901.10008v2
PDF	http://arxiv.org/pdf/1901.10008v2.pdf
PWC	https://paperswithcode.com/paper/the-ooo-vliw-jit-compiler-for-gpu-inference
Repo
Framework