IJCAI 2020 论文列表

Reinforcement learning agents usually learn from scratch, which requires a large number of interactions with the environment. This is quite different from the learning process of human. When faced with a new task, human naturally have the common sense and use the prior knowledge to derive an initial policy and guide the learning process afterwards. Although the prior knowledge may be not fully applicable to the new task, the learning process is significantly sped up since the initial policy ensures a quick-start of learning and intermediate guidance allows to avoid unnecessary exploration. Taking this inspiration, we propose knowledge guided policy network (KoGuN), a novel framework that combines human prior suboptimal knowledge with reinforcement learning. Our framework consists of a fuzzy rule controller to represent human knowledge and a refine module to finetune suboptimal prior knowledge. The proposed framework is end-to-end and can be combined with existing policy-based reinforcement learning algorithm. We conduct experiments on several control tasks. The empirical results show that our approach, which combines suboptimal human knowledge and RL, achieves significant improvement on learning efficiency of flat RL algorithms, even with very low-performance human prior knowledge.

Prior domain knowledge can greatly help to learn generative models. However, it is often too costly to hard-code prior knowledge as a specific model architecture, so we often have to use general-purpose models. In this paper, we propose a method to incorporate prior knowledge of feature relations into the learning of general-purpose generative models. To this end, we formulate a regularizer that makes the marginals of a generative model to follow prescribed relative dependence of features. It can be incorporated into off-the-shelf learning methods of many generative models, including variational autoencoders and generative adversarial networks, as its gradients can be computed using standard backpropagation techniques. We show the effectiveness of the proposed method with experiments on multiple types of datasets and generative models.

Knowledge graphs store facts using relations between two entities. In this work, we address the question of link prediction in knowledge hypergraphs where relations are defined on any number of entities. While techniques exist (such as reification) that convert non-binary relations into binary ones, we show that current embedding-based methods for knowledge graph completion do not work well out of the box for knowledge graphs obtained through these techniques. To overcome this, we introduce HSimplE and HypE, two embedding-based methods that work directly with knowledge hypergraphs. In both models, the prediction is a function of the relation embedding, the entity embeddings and their corresponding positions in the relation. We also develop public datasets, benchmarks and baselines for hypergraph prediction and show experimentally that the proposed models are more effective than the baselines.

Knowledge graphs (KGs) store much structured information on various entities, many of which are not covered by the parallel sentence pairs of neural machine translation (NMT). To improve the translation quality of these entities, in this paper we propose a novel KGs enhanced NMT method. Specifically, we first induce the new translation results of these entities by transforming the source and target KGs into a unified semantic space. We then generate adequate pseudo parallel sentence pairs that contain these induced entity pairs. Finally, NMT model is jointly trained by the original and pseudo sentence pairs. The extensive experiments on Chinese-to-English and Englishto-Japanese translation tasks demonstrate that our method significantly outperforms the strong baseline models in translation quality, especially in handling the induced entities.

Identifying causal relations of events is a crucial language understanding task. Despite many efforts for this task, existing methods lack the ability to adopt background knowledge, and they typically generalize poorly to new, previously unseen data. In this paper, we present a new method for event causality identification, aiming to address limitations of previous methods. On the one hand, our model can leverage external knowledge for reasoning, which can greatly enrich the representation of events; On the other hand, our model can mine event-agnostic, context-specific patterns, via a mechanism called event mention masking generalization, which can greatly enhance the ability of our model to handle new, previously unseen cases. In experiments, we evaluate our model on three benchmark datasets and show our model outperforms previous methods by a significant margin. Moreover, we perform 1) cross-topic adaptation, 2) exploiting unseen predicates, and 3) cross-task adaptation to evaluate the generalization ability of our model. Experimental results show that our model demonstrates a definite advantage over previous methods.

Logical systems containing knowledge and know-how modalities have been investigated in several recent works. Independently, epistemic modal logics in which every knowledge modality is labeled with a degree of uncertainty have been proposed. This article combines these two research lines by introducing a bimodal logic containing knowledge and know-how modalities, both labeled with a degree of uncertainty. The main technical results are soundness, completeness, and incompleteness of the proposed logical system with respect to two classes of semantics.

We examine the problem of assigning plots of land to prospective buyers who prefer living next to their friends. In this setting, each agent's utility depends on the plot she receives and the identities of the agents who receive the adjacent plots. We are interested in mechanisms without money that guarantee truthful reporting of both land values and friendships, as well as Pareto optimality and computational efficiency. We explore several modifications of the Random Serial Dictatorship (RSD) mechanism, and identify one that performs well according to these criteria, We also study the expected social welfare of the assignments produced by our mechanisms when agents' values for the land plots are binary; it turns out that we can achieve good approximations to the optimal social welfare, but only if the agents value the friendships highly.

This paper proposed a new interaction paradigm in the virtual reality (VR) environments, which consists of a virtual mirror or window projected onto a virtual surface, representing the correct perspective geometry of a mirror or window reflecting the real world. This technique can be applied to various videos, live streaming apps, augmented and virtual reality settings to provide an interactive and immersive user experience. To support such a perspective-accurate representation, we implemented computer vision algorithms for feature detection and correspondence matching. To constrain the solutions, we incorporated an automatically tuning scaling factor upon the homography transform matrix such that each image frame follows a smooth transition with the user in sight. The system is a real-time rendering framework where users can engage their real-life presence with the virtual space.

Drug-drug interaction (DDI) prediction is a challenging problem in pharmacology and clinical application, and effectively identifying potential DDIs during clinical trials is critical for patients and society. Most of existing computational models with AI techniques often concentrate on integrating multiple data sources and combining popular embedding methods together. Yet, researchers pay less attention to the potential correlations between drug and other entities such as targets and genes. Moreover, recent studies also adopted knowledge graph (KG) for DDI prediction. Yet, this line of methods learn node latent embedding directly, but they are limited in obtaining the rich neighborhood information of each entity in the KG. To address the above limitations, we propose an end-to-end framework, called Knowledge Graph Neural Network (KGNN), to resolve the DDI prediction. Our framework can effectively capture drug and its potential neighborhoods by mining their associated relations in KG. To extract both high-order structures and semantic relations of the KG, we learn from the neighborhoods for each entity in the KG as their local receptive, and then integrate neighborhood information with bias from representation of the current entity. This way, the receptive field can be naturally extended to multiple hops away to model high-order topological information and to obtain drugs potential long-distance correlations. We have implemented our method and conducted experiments based on several widely-used datasets. Empirical results show that KGNN outperforms the classic and state-of-the-art models.

For single-channel speech enhancement, both time-domain and time-frequency-domain methods have their respective pros and cons. In this paper, we present a cross-domain framework named TFT-Net, which takes time-frequency spectrogram as input and produces time-domain waveform as output. Such a framework takes advantage of the knowledge we have about spectrogram and avoids some of the drawbacks that T-F-domain methods have been suffering from. In TFT-Net, we design an innovative dual-path attention block (DAB) to fully exploit correlations along the time and frequency axes. We further discover that a sample-independent DAB (SDAB) achieves a good tradeoff between enhanced speech quality and complexity. Ablation studies show that both the cross-domain design and the SDAB block bring large performance gain. When logarithmic MSE is used as the training criteria, TFT-Net achieves the highest SDR and SSNR among state-of-the-art methods on two major speech enhancement benchmarks.

In this paper, we explore to learn representations of legislation and legislator for the prediction of roll call results. The most popular approach for this topic is named the ideal point model that relies on historical voting information for representation learning of legislators. It largely ignores the context information of the legislative data. We, therefore, propose to incorporate context information to learn dense representations for both legislators and legislation. For legislators, we incorporate relations among them via graph convolutional neural networks (GCN) for their representation learning. For legislation, we utilize its narrative description via recurrent neural networks (RNN) for representation learning. In order to align two kinds of representations in the same vector space, we introduce a triplet loss for the joint training. Experimental results on a self-constructed dataset show the effectiveness of our model for roll call results prediction compared to some state-of-the-art baselines.

Domain adaptation (DA) has achieved a resounding success to learn a good classifier by leveraging labeled data from a source domain to adapt to an unlabeled target domain. However, in a general setting when the target domain contains classes that are never observed in the source domain, namely in Open Set Domain Adaptation (OSDA), existing DA methods failed to work because of the interference of the extra unknown classes. This is a much more challenging problem, since it can easily result in negative transfer due to the mismatch between the unknown and known classes. Existing researches are susceptible to misclassification when target domain unknown samples in the feature space distributed near the decision boundary learned from the labeled source domain. To overcome this, we propose Joint Partial Optimal Transport (JPOT), fully utilizing information of not only the labeled source domain but also the discriminative representation of unknown class in the target domain. The proposed joint discriminative prototypical compactness loss can not only achieve intra-class compactness and inter-class separability, but also estimate the mean and variance of the unknown class through backpropagation, which remains intractable for previous methods due to the blindness about the structure of the unknown classes. To our best knowledge, this is the first optimal transport model for OSDA. Extensive experiments demonstrate that our proposed model can significantly boost the performance of open set domain adaptation on standard DA datasets.

Three-dimensional (3D) object classification is widely involved in various computer vision applications, e.g., autonomous driving, simultaneous localization and mapping, which has attracted lots of attention in the committee. However, solving 3D object classification by directly employing the 3D convolutional neural networks (CNNs) generally suffers from high computational cost. Besides, existing view-based methods cannot better explore the content relationships between views. To this end, this work proposes a novel multi-view framework by jointly using multiple 2D-CNNs to capture discriminative information with relationships as well as a new multi-view loss fusion strategy, in an end-to-end manner. Specifically, we utilize multiple 2D views of a 3D object as input and integrate the intra-view and inter-view information of each view through the view-specific 2D-CNN and a series of modules (outer product, view pair pooling, 1D convolution, and fully connected transformation). Furthermore, we design a novel view ensemble mechanism that selects several discriminative and informative views to jointly infer the category of a 3D object. Extensive experiments demonstrate that the proposed method is able to outperform current state-of-the-art methods on 3D object classification. More importantly, this work provides a new way to improve 3D object classification from the perspective of fully utilizing well-established 2D-CNNs.

Existing deep learning-based image de-blocking methods use only pixel-level loss functions to guide network training. The JPEG compression factor, which reflects the degradation degree, has not been fully utilized. However, due to the non-differentiability, the compression factor cannot be directly utilized to train deep networks. To solve this problem, we propose compression quality ranker-guided networks for this specific JPEG artifacts removal. We first design a quality ranker to measure the compression degree, which is highly correlated with the JPEG quality. Based on this differentiable ranker, we then propose one quality-related loss and one feature matching loss to guide de-blocking and perceptual quality optimization. In addition, we utilize dilated convolutions to extract multi-scale features, which enables our single model to handle multiple compression quality factors. Our method can implicitly use the information contained in the compression factors to produce better results. Experiments demonstrate that our model can achieve comparable or even better performance in both quantitative and qualitative measurements.

Motivated by the capability of Generative Adversarial Network on exploring the latent semantic space and capturing semantic variations in the data distribution, adversarial learning has been adopted in network embedding to improve the robustness. However, this important ability is lost in existing adversarially regularized network embedding methods, because their embedding results are directly compared to the samples drawn from perturbation (Gaussian) distribution without any rectification from real data. To overcome this vital issue, a novel Joint Adversarial Network Embedding (JANE) framework is proposed to jointly distinguish the real and fake combinations of the embeddings, topology information and node features. JANE contains three pluggable components, Embedding module, Generator module and Discriminator module. The overall objective function of JANE is defined in a min-max form, which can be optimized via alternating stochastic gradient. Extensive experiments demonstrate the remarkable superiority of the proposed JANE on link prediction (3% gains in both AUC and AP) and node clustering (5% gain in F1 score).

Hierarchical reinforcement learning (HRL) is a promising approach to solve tasks with long time horizons and sparse rewards. It is often implemented as a high-level policy assigning subgoals to a low-level policy. However, it suffers the high-level non-stationarity problem since the low-level policy is constantly changing. The non-stationarity also leads to the data efficiency problem: policies need more data at non-stationary states to stabilize training. To address these issues, we propose a novel HRL method: Interactive Influence-based Hierarchical Reinforcement Learning (I^2HRL). First, inspired by agent modeling, we enable the interaction between the low-level and high-level policies to stabilize the high-level policy training. The high-level policy makes decisions conditioned on the received low-level policy representation as well as the state of the environment. Second, we furthermore stabilize the high-level policy via an information-theoretic regularization with minimal dependence on the changing low-level policy. Third, we propose the influence-based exploration to more frequently visit the non-stationary states where more transition data is needed. We experimentally validate the effectiveness of the proposed solution in several tasks in MuJoCo domains by demonstrating that our approach can significantly boost the learning performance and accelerate learning compared with state-of-the-art HRL methods.

Conflict-Based Search (CBS) is a leading algorithm for optimal Multi-Agent Path Finding (MAPF). CBS variants typically compute MAPF solutions using some form of A* search. However, they often do so under strict time limits so as to avoid exhausting the available memory. In this paper, we present IDCBS, an iterative-deepening variant of CBS which can be executed without exhausting the memory and without strict time limits. IDCBS can be substantially faster than CBS due to incremental methods that it uses when processing CBS nodes.

The residual network is now one of the most effective structures in deep learning, which utilizes the skip connections to “guarantee" the performance will not get worse. However, the non-convexity of the neural network makes it unclear whether the skip connections do provably improve the learning ability since the nonlinearity may create many local minima. In some previous works [Freeman and Bruna, 2016], it is shown that despite the non-convexity, the loss landscape of the two-layer ReLU network has good properties when the number m of hidden nodes is very large. In this paper, we follow this line to study the topology (sub-level sets) of the loss landscape of deep ReLU neural networks with a skip connection and theoretically prove that the skip connection network inherits the good properties of the two-layer network and skip connections can help to control the connectedness of the sub-level sets, such that any local minima worse than the global minima of some two-layer ReLU network will be very “shallow". The “depth" of these local minima are at most O(m^(η-1)/n), where n is the input dimension, η<1. This provides a theoretical explanation for the effectiveness of the skip connection in deep learning.

A major task of sports analytics is to rank players based on the impact of their actions. Recent methods have applied reinforcement learning (RL) to assess the value of actions from a learned action value or Q-function. A fundamental challenge for estimating action values is that explicit reward signals (goals) are very sparse in many team sports, such as ice hockey and soccer. This paper combines Q-function learning with inverse reinforcement learning (IRL) to provide a novel player ranking method. We treat professional play as expert demonstrations for learning an implicit reward function. Our method alternates single-agent IRL to learn a reward function for multiple agents; we provide a theoretical justification for this procedure. Knowledge transfer is used to combine learned rewards and observed rewards from goals. Empirical evaluation, based on 4.5M play-by-play events in the National Hockey League (NHL), indicates that player ranking using the learned rewards achieves high correlations with standard success measures and temporal consistency throughout a season.