October 21, 2019

3364 words 16 mins read

Paper Group AWR 70

Unbiased deep solvers for parametric PDEs. StarCraft Micromanagement with Reinforcement Learning and Curriculum Transfer Learning. SupportNet: solving catastrophic forgetting in class incremental learning with support data. Multi-Pointer Co-Attention Networks for Recommendation. Context-Aware Visual Policy Network for Sequence-Level Image Captionin …

Unbiased deep solvers for parametric PDEs


Title	Unbiased deep solvers for parametric PDEs
Authors	Marc Sabate Vidales, David Siska, Lukasz Szpruch
Abstract	We develop several deep learning algorithms for approximating families of parametric PDE solutions. The proposed algorithms approximate solutions together with their gradients, which in the context of mathematical finance means that the derivative prices and hedging strategies are computed simulatenously. Having approximated the gradient of the solution one can combine it with a Monte-Carlo simulation to remove the bias in the deep network approximation of the PDE solution (derivative price). This is achieved by leveraging the Martingale Representation Theorem and combining the Monte Carlo simulation with the neural network. The resulting algorithm is robust with respect to quality of the neural network approximation and consequently can be used as a black-box in case only limited a priori information about the underlying problem is available. We believe this is important as neural network based algorithms often require fair amount of tuning to produce satisfactory results. The methods are empirically shown to work for high-dimensional problems (e.g. 100 dimensions). We provide diagnostics that shed light on appropriate network architectures.
Tasks
Published	2018-10-11
URL	https://arxiv.org/abs/1810.05094v2
PDF	https://arxiv.org/pdf/1810.05094v2.pdf
PWC	https://paperswithcode.com/paper/martingale-functional-control-variates-via
Repo	https://github.com/marcsv87/Deep-PDE-Solvers
Framework	pytorch

StarCraft Micromanagement with Reinforcement Learning and Curriculum Transfer Learning


Title	StarCraft Micromanagement with Reinforcement Learning and Curriculum Transfer Learning
Authors	Kun Shao, Yuanheng Zhu, Dongbin Zhao
Abstract	Real-time strategy games have been an important field of game artificial intelligence in recent years. This paper presents a reinforcement learning and curriculum transfer learning method to control multiple units in StarCraft micromanagement. We define an efficient state representation, which breaks down the complexity caused by the large state space in the game environment. Then a parameter sharing multi-agent gradientdescent Sarsa({\lambda}) (PS-MAGDS) algorithm is proposed to train the units. The learning policy is shared among our units to encourage cooperative behaviors. We use a neural network as a function approximator to estimate the action-value function, and propose a reward function to help units balance their move and attack. In addition, a transfer learning method is used to extend our model to more difficult scenarios, which accelerates the training process and improves the learning performance. In small scale scenarios, our units successfully learn to combat and defeat the built-in AI with 100% win rates. In large scale scenarios, curriculum transfer learning method is used to progressively train a group of units, and shows superior performance over some baseline methods in target scenarios. With reinforcement learning and curriculum transfer learning, our units are able to learn appropriate strategies in StarCraft micromanagement scenarios.
Tasks	Real-Time Strategy Games, Starcraft, Transfer Learning
Published	2018-04-03
URL	http://arxiv.org/abs/1804.00810v1
PDF	http://arxiv.org/pdf/1804.00810v1.pdf
PWC	https://paperswithcode.com/paper/starcraft-micromanagement-with-reinforcement
Repo	https://github.com/DRL-CASIA/Game-AI
Framework	none

SupportNet: solving catastrophic forgetting in class incremental learning with support data


Title	SupportNet: solving catastrophic forgetting in class incremental learning with support data
Authors	Yu Li, Zhongxiao Li, Lizhong Ding, Yijie Pan, Chao Huang, Yuhui Hu, Wei Chen, Xin Gao
Abstract	A plain well-trained deep learning model often does not have the ability to learn new knowledge without forgetting the previously learned knowledge, which is known as catastrophic forgetting. Here we propose a novel method, SupportNet, to efficiently and effectively solve the catastrophic forgetting problem in the class incremental learning scenario. SupportNet combines the strength of deep learning and support vector machine (SVM), where SVM is used to identify the support data from the old data, which are fed to the deep learning model together with the new data for further training so that the model can review the essential information of the old data when learning the new information. Two powerful consolidation regularizers are applied to stabilize the learned representation and ensure the robustness of the learned model. We validate our method with comprehensive experiments on various tasks, which show that SupportNet drastically outperforms the state-of-the-art incremental learning methods and even reaches similar performance as the deep learning model trained from scratch on both old and new data. Our program is accessible at: https://github.com/lykaust15/SupportNet
Tasks
Published	2018-06-08
URL	http://arxiv.org/abs/1806.02942v3
PDF	http://arxiv.org/pdf/1806.02942v3.pdf
PWC	https://paperswithcode.com/paper/supportnet-solving-catastrophic-forgetting-in
Repo	https://github.com/lykaust15/SupportNet
Framework	tf

Multi-Pointer Co-Attention Networks for Recommendation


Title	Multi-Pointer Co-Attention Networks for Recommendation
Authors	Yi Tay, Luu Anh Tuan, Siu Cheung Hui
Abstract	Many recent state-of-the-art recommender systems such as D-ATT, TransNet and DeepCoNN exploit reviews for representation learning. This paper proposes a new neural architecture for recommendation with reviews. Our model operates on a multi-hierarchical paradigm and is based on the intuition that not all reviews are created equal, i.e., only a select few are important. The importance, however, should be dynamically inferred depending on the current target. To this end, we propose a review-by-review pointer-based learning scheme that extracts important reviews, subsequently matching them in a word-by-word fashion. This enables not only the most informative reviews to be utilized for prediction but also a deeper word-level interaction. Our pointer-based method operates with a novel gumbel-softmax based pointer mechanism that enables the incorporation of discrete vectors within differentiable neural architectures. Our pointer mechanism is co-attentive in nature, learning pointers which are co-dependent on user-item relationships. Finally, we propose a multi-pointer learning scheme that learns to combine multiple views of interactions between user and item. Overall, we demonstrate the effectiveness of our proposed model via extensive experiments on \textbf{24} benchmark datasets from Amazon and Yelp. Empirical results show that our approach significantly outperforms existing state-of-the-art, with up to 19% and 71% relative improvement when compared to TransNet and DeepCoNN respectively. We study the behavior of our multi-pointer learning mechanism, shedding light on evidence aggregation patterns in review-based recommender systems.
Tasks	Recommendation Systems, Representation Learning
Published	2018-01-28
URL	http://arxiv.org/abs/1801.09251v2
PDF	http://arxiv.org/pdf/1801.09251v2.pdf
PWC	https://paperswithcode.com/paper/multi-pointer-co-attention-networks-for
Repo	https://github.com/vanzytay/KDD2018_MPCN
Framework	tf

Context-Aware Visual Policy Network for Sequence-Level Image Captioning


Title	Context-Aware Visual Policy Network for Sequence-Level Image Captioning
Authors	Daqing Liu, Zheng-Jun Zha, Hanwang Zhang, Yongdong Zhang, Feng Wu
Abstract	Many vision-language tasks can be reduced to the problem of sequence prediction for natural language output. In particular, recent advances in image captioning use deep reinforcement learning (RL) to alleviate the “exposure bias” during training: ground-truth subsequence is exposed in every step prediction, which introduces bias in test when only predicted subsequence is seen. However, existing RL-based image captioning methods only focus on the language policy while not the visual policy (e.g., visual attention), and thus fail to capture the visual context that are crucial for compositional reasoning such as visual relationships (e.g., “man riding horse”) and comparisons (e.g., “smaller cat”). To fill the gap, we propose a Context-Aware Visual Policy network (CAVP) for sequence-level image captioning. At every time step, CAVP explicitly accounts for the previous visual attentions as the context, and then decides whether the context is helpful for the current word generation given the current visual attention. Compared against traditional visual attention that only fixes a single image region at every step, CAVP can attend to complex visual compositions over time. The whole image captioning model — CAVP and its subsequent language policy network — can be efficiently optimized end-to-end by using an actor-critic policy gradient method with respect to any caption evaluation metric. We demonstrate the effectiveness of CAVP by state-of-the-art performances on MS-COCO offline split and online server, using various metrics and sensible visualizations of qualitative visual context. The code is available at https://github.com/daqingliu/CAVP
Tasks	Image Captioning
Published	2018-08-16
URL	http://arxiv.org/abs/1808.05864v3
PDF	http://arxiv.org/pdf/1808.05864v3.pdf
PWC	https://paperswithcode.com/paper/context-aware-visual-policy-network-for
Repo	https://github.com/daqingliu/CAVP
Framework	pytorch

Decoupling Structure and Lexicon for Zero-Shot Semantic Parsing


Title	Decoupling Structure and Lexicon for Zero-Shot Semantic Parsing
Authors	Jonathan Herzig, Jonathan Berant
Abstract	Building a semantic parser quickly in a new domain is a fundamental challenge for conversational interfaces, as current semantic parsers require expensive supervision and lack the ability to generalize to new domains. In this paper, we introduce a zero-shot approach to semantic parsing that can parse utterances in unseen domains while only being trained on examples in other source domains. First, we map an utterance to an abstract, domain-independent, logical form that represents the structure of the logical form, but contains slots instead of KB constants. Then, we replace slots with KB constants via lexical alignment scores and global inference. Our model reaches an average accuracy of 53.4% on 7 domains in the Overnight dataset, substantially better than other zero-shot baselines, and performs as good as a parser trained on over 30% of the target domain examples.
Tasks	Semantic Parsing
Published	2018-04-21
URL	http://arxiv.org/abs/1804.07918v2
PDF	http://arxiv.org/pdf/1804.07918v2.pdf
PWC	https://paperswithcode.com/paper/decoupling-structure-and-lexicon-for-zero
Repo	https://github.com/jonathanherzig/zero-shot-semantic-parsing
Framework	none

Encoder-Decoder with Atrous Separable Convolution for Semantic Image Segmentation


Title	Encoder-Decoder with Atrous Separable Convolution for Semantic Image Segmentation
Authors	Liang-Chieh Chen, Yukun Zhu, George Papandreou, Florian Schroff, Hartwig Adam
Abstract	Spatial pyramid pooling module or encode-decoder structure are used in deep neural networks for semantic segmentation task. The former networks are able to encode multi-scale contextual information by probing the incoming features with filters or pooling operations at multiple rates and multiple effective fields-of-view, while the latter networks can capture sharper object boundaries by gradually recovering the spatial information. In this work, we propose to combine the advantages from both methods. Specifically, our proposed model, DeepLabv3+, extends DeepLabv3 by adding a simple yet effective decoder module to refine the segmentation results especially along object boundaries. We further explore the Xception model and apply the depthwise separable convolution to both Atrous Spatial Pyramid Pooling and decoder modules, resulting in a faster and stronger encoder-decoder network. We demonstrate the effectiveness of the proposed model on PASCAL VOC 2012 and Cityscapes datasets, achieving the test set performance of 89.0% and 82.1% without any post-processing. Our paper is accompanied with a publicly available reference implementation of the proposed models in Tensorflow at \url{https://github.com/tensorflow/models/tree/master/research/deeplab}.
Tasks	Image Classification, Lesion Segmentation, Semantic Segmentation
Published	2018-02-07
URL	http://arxiv.org/abs/1802.02611v3
PDF	http://arxiv.org/pdf/1802.02611v3.pdf
PWC	https://paperswithcode.com/paper/encoder-decoder-with-atrous-separable
Repo	https://github.com/shenshutao/image_segmentation
Framework	tf

Semantic White Balance: Semantic Color Constancy Using Convolutional Neural Network


Title	Semantic White Balance: Semantic Color Constancy Using Convolutional Neural Network
Authors	Mahmoud Afifi
Abstract	The goal of computational color constancy is to preserve the perceptive colors of objects under different lighting conditions by removing the effect of color casts caused by the scene’s illumination. With the rapid development of deep learning based techniques, significant progress has been made in image semantic segmentation. In this work, we exploit the semantic information together with the color and spatial information of the input image in order to remove color casts. We train a convolutional neural network (CNN) model that learns to estimate the illuminant color and gamma correction parameters based on the semantic information of the given image. Experimental results show that feeding the CNN with the semantic information leads to a significant improvement in the results by reducing the error by more than 40%.
Tasks	Color Constancy, Semantic Segmentation
Published	2018-02-01
URL	https://arxiv.org/abs/1802.00153v5
PDF	https://arxiv.org/pdf/1802.00153v5.pdf
PWC	https://paperswithcode.com/paper/semantic-white-balance-semantic-color
Repo	https://github.com/mahmoudnafifi/Semantic-Color-Constancy-Using-CNN
Framework	tf

DPASF: A Flink Library for Streaming Data preprocessing


Title	DPASF: A Flink Library for Streaming Data preprocessing
Authors	Alejandro Alcalde-Barros, Diego García-Gil, Salvador García, Francisco Herrera
Abstract	Data preprocessing techniques are devoted to correct or alleviate errors in data. Discretization and feature selection are two of the most extended data preprocessing techniques. Although we can find many proposals for static Big Data preprocessing, there is little research devoted to the continuous Big Data problem. Apache Flink is a recent and novel Big Data framework, following the MapReduce paradigm, focused on distributed stream and batch data processing. In this paper we propose a data stream library for Big Data preprocessing, named DPASF, under Apache Flink. We have implemented six of the most popular data preprocessing algorithms, three for discretization and the rest for feature selection. The algorithms have been tested using two Big Data datasets. Experimental results show that preprocessing can not only reduce the size of the data, but to maintain or even improve the original accuracy in a short time. DPASF contains useful algorithms when dealing with Big Data data streams. The preprocessing algorithms included in the library are able to tackle Big Datasets efficiently and to correct imperfections in the data.
Tasks	Feature Selection
Published	2018-10-14
URL	http://arxiv.org/abs/1810.06021v1
PDF	http://arxiv.org/pdf/1810.06021v1.pdf
PWC	https://paperswithcode.com/paper/dpasf-a-flink-library-for-streaming-data
Repo	https://github.com/elbaulp/dpasf
Framework	none

Unified Hypersphere Embedding for Speaker Recognition


Title	Unified Hypersphere Embedding for Speaker Recognition
Authors	Mahdi Hajibabaei, Dengxin Dai
Abstract	Incremental improvements in accuracy of Convolutional Neural Networks are usually achieved through use of deeper and more complex models trained on larger datasets. However, enlarging dataset and models increases the computation and storage costs and cannot be done indefinitely. In this work, we seek to improve the identification and verification accuracy of a text-independent speaker recognition system without use of extra data or deeper and more complex models by augmenting the training and testing data, finding the optimal dimensionality of embedding space and use of more discriminative loss functions. Results of experiments on VoxCeleb dataset suggest that: (i) Simple repetition and random time-reversion of utterances can reduce prediction errors by up to 18%. (ii) Lower dimensional embeddings are more suitable for verification. (iii) Use of proposed logistic margin loss function leads to unified embeddings with state-of-the-art identification and competitive verification accuracies.
Tasks	Speaker Recognition, Text-Independent Speaker Recognition
Published	2018-07-22
URL	http://arxiv.org/abs/1807.08312v1
PDF	http://arxiv.org/pdf/1807.08312v1.pdf
PWC	https://paperswithcode.com/paper/unified-hypersphere-embedding-for-speaker
Repo	https://github.com/MahdiHajibabaei/unified-embedding
Framework	caffe2

How To Backdoor Federated Learning


Title	How To Backdoor Federated Learning
Authors	Eugene Bagdasaryan, Andreas Veit, Yiqing Hua, Deborah Estrin, Vitaly Shmatikov
Abstract	Federated learning enables thousands of participants to construct a deep learning model without sharing their private training data with each other. For example, multiple smartphones can jointly train a next-word predictor for keyboards without revealing what individual users type. We demonstrate that any participant in federated learning can introduce hidden backdoor functionality into the joint global model, e.g., to ensure that an image classifier assigns an attacker-chosen label to images with certain features, or that a word predictor completes certain sentences with an attacker-chosen word. We design and evaluate a new model-poisoning methodology based on model replacement. An attacker selected in a single round of federated learning can cause the global model to immediately reach 100% accuracy on the backdoor task. We evaluate the attack under different assumptions for the standard federated-learning tasks and show that it greatly outperforms data poisoning. Our generic constrain-and-scale technique also evades anomaly detection-based defenses by incorporating the evasion into the attacker’s loss function during training.
Tasks	Anomaly Detection, data poisoning
Published	2018-07-02
URL	https://arxiv.org/abs/1807.00459v3
PDF	https://arxiv.org/pdf/1807.00459v3.pdf
PWC	https://paperswithcode.com/paper/how-to-backdoor-federated-learning
Repo	https://github.com/ebagdasa/backdoor_federated_learning
Framework	pytorch

Polisis: Automated Analysis and Presentation of Privacy Policies Using Deep Learning


Title	Polisis: Automated Analysis and Presentation of Privacy Policies Using Deep Learning
Authors	Hamza Harkous, Kassem Fawaz, Rémi Lebret, Florian Schaub, Kang G. Shin, Karl Aberer
Abstract	Privacy policies are the primary channel through which companies inform users about their data collection and sharing practices. These policies are often long and difficult to comprehend. Short notices based on information extracted from privacy policies have been shown to be useful but face a significant scalability hurdle, given the number of policies and their evolution over time. Companies, users, researchers, and regulators still lack usable and scalable tools to cope with the breadth and depth of privacy policies. To address these hurdles, we propose an automated framework for privacy policy analysis (Polisis). It enables scalable, dynamic, and multi-dimensional queries on natural language privacy policies. At the core of Polisis is a privacy-centric language model, built with 130K privacy policies, and a novel hierarchy of neural-network classifiers that accounts for both high-level aspects and fine-grained details of privacy practices. We demonstrate Polisis’ modularity and utility with two applications supporting structured and free-form querying. The structured querying application is the automated assignment of privacy icons from privacy policies. With Polisis, we can achieve an accuracy of 88.4% on this task. The second application, PriBot, is the first freeform question-answering system for privacy policies. We show that PriBot can produce a correct answer among its top-3 results for 82% of the test questions. Using an MTurk user study with 700 participants, we show that at least one of PriBot’s top-3 answers is relevant to users for 89% of the test questions.
Tasks	Language Modelling, Question Answering
Published	2018-02-07
URL	http://arxiv.org/abs/1802.02561v2
PDF	http://arxiv.org/pdf/1802.02561v2.pdf
PWC	https://paperswithcode.com/paper/polisis-automated-analysis-and-presentation
Repo	https://github.com/wi-pi/GDPR
Framework	none

A Systematic DNN Weight Pruning Framework using Alternating Direction Method of Multipliers


Title	A Systematic DNN Weight Pruning Framework using Alternating Direction Method of Multipliers
Authors	Tianyun Zhang, Shaokai Ye, Kaiqi Zhang, Jian Tang, Wujie Wen, Makan Fardad, Yanzhi Wang
Abstract	Weight pruning methods for deep neural networks (DNNs) have been investigated recently, but prior work in this area is mainly heuristic, iterative pruning, thereby lacking guarantees on the weight reduction ratio and convergence time. To mitigate these limitations, we present a systematic weight pruning framework of DNNs using the alternating direction method of multipliers (ADMM). We first formulate the weight pruning problem of DNNs as a nonconvex optimization problem with combinatorial constraints specifying the sparsity requirements, and then adopt the ADMM framework for systematic weight pruning. By using ADMM, the original nonconvex optimization problem is decomposed into two subproblems that are solved iteratively. One of these subproblems can be solved using stochastic gradient descent, the other can be solved analytically. Besides, our method achieves a fast convergence rate. The weight pruning results are very promising and consistently outperform the prior work. On the LeNet-5 model for the MNIST data set, we achieve 71.2 times weight reduction without accuracy loss. On the AlexNet model for the ImageNet data set, we achieve 21 times weight reduction without accuracy loss. When we focus on the convolutional layer pruning for computation reductions, we can reduce the total computation by five times compared with the prior work (achieving a total of 13.4 times weight reduction in convolutional layers). Our models and codes are released at https://github.com/KaiqiZhang/admm-pruning
Tasks
Published	2018-04-10
URL	http://arxiv.org/abs/1804.03294v3
PDF	http://arxiv.org/pdf/1804.03294v3.pdf
PWC	https://paperswithcode.com/paper/a-systematic-dnn-weight-pruning-framework
Repo	https://github.com/KaiqiZhang/caffe-admm
Framework	none

Message Passing Graph Kernels


Title	Message Passing Graph Kernels
Authors	Giannis Nikolentzos, Michalis Vazirgiannis
Abstract	Graph kernels have recently emerged as a promising approach for tackling the graph similarity and learning tasks at the same time. In this paper, we propose a general framework for designing graph kernels. The proposed framework capitalizes on the well-known message passing scheme on graphs. The kernels derived from the framework consist of two components. The first component is a kernel between vertices, while the second component is a kernel between graphs. The main idea behind the proposed framework is that the representations of the vertices are implicitly updated using an iterative procedure. Then, these representations serve as the building blocks of a kernel that compares pairs of graphs. We derive four instances of the proposed framework, and show through extensive experiments that these instances are competitive with state-of-the-art methods in various tasks.
Tasks	Graph Similarity
Published	2018-08-07
URL	http://arxiv.org/abs/1808.02510v1
PDF	http://arxiv.org/pdf/1808.02510v1.pdf
PWC	https://paperswithcode.com/paper/message-passing-graph-kernels
Repo	https://github.com/giannisnik/message_passing_graph_kernels
Framework	none

Triple Trustworthiness Measurement for Knowledge Graph


Title	Triple Trustworthiness Measurement for Knowledge Graph
Authors	Shengbin Jia, Yang Xiang, Xiaojun Chen
Abstract	The Knowledge graph (KG) uses the triples to describe the facts in the real world. It has been widely used in intelligent analysis and applications. However, possible noises and conflicts are inevitably introduced in the process of constructing. And the KG based tasks or applications assume that the knowledge in the KG is completely correct and inevitably bring about potential deviations. In this paper, we establish a knowledge graph triple trustworthiness measurement model that quantify their semantic correctness and the true degree of the facts expressed. The model is a crisscrossing neural network structure. It synthesizes the internal semantic information in the triples and the global inference information of the KG to achieve the trustworthiness measurement and fusion in the three levels of entity level, relationship level, and KG global level. We analyzed the validity of the model output confidence values, and conducted experiments in the real-world dataset FB15K (from Freebase) for the knowledge graph error detection task. The experimental results showed that compared with other models, our model achieved significant and consistent improvements.
Tasks
Published	2018-09-25
URL	http://arxiv.org/abs/1809.09414v3
PDF	http://arxiv.org/pdf/1809.09414v3.pdf
PWC	https://paperswithcode.com/paper/triple-trustworthiness-measurement-for
Repo	https://github.com/TJUNLP/TTMF
Framework	none