IJCAI 2020 Papers

Recent proliferation of cryptocurrencies that allow for pseudo-anonymous transactions has resulted in a spike of various e-crime activities and, particularly, cryptocurrency payments in hacking attacks demanding ransom by encrypting sensitive user data. Currently, most hackers use Bitcoin for payments, and existing ransomware detection tools depend only on a couple of heuristics and/or tedious data gathering steps. By capitalizing on the recent advances in Topological Data Analysis, we propose a novel efficient and tractable framework to automatically predict new ransomware transactions in a ransomware family, given only limited records of past transactions. Moreover, our new methodology exhibits high utility to detect emergence of new ransomware families, that is, detecting ransomware with no past records of transactions.

Constraint Satisfaction Problems (CSPs) are typically solved with Generalized Arc Consistency (GAC). A general CSP can also be encoded into a binary CSP and solved with Arc Consistency (AC). The well-known Hidden Variable Encoding (HVE) is still a state-of-the-art binary encoding for solving CSPs. We propose a new binary encoding, called Bipartite Encoding (BE) which uses the idea of partitioning constraints. A BE encoded CSP can achieve a higher level of consistency than GAC on the original CSP. We give an algorithm for creating compact bipartite encoding for non-binary CSPs. We present a AC propagator on the binary constraints from BE exploiting their special structure. Experiments on a large set of non-binary CSP benchmarks with table constraints using the Wdeg, Activity and Impact heuristics show that BE with our AC propagator can outperform existing state-of-the-art GAC algorithms (CT, STRbit) and binary encodings (HVE with HTAC).

Positive-confidence (Pconf) classification [Ishida et al., 2018] is a promising weakly-supervised learning method which trains a binary classifier only from positive data equipped with confidence. However, in practice, the confidence may be skewed by bias arising in an annotation process. The Pconf classifier cannot be properly learned with skewed confidence, and consequently, the classification performance might be deteriorated. In this paper, we introduce the parameterized model of the skewed confidence, and propose the method for selecting the hyperparameter which cancels out the negative impact of the skewed confidence under the assumption that we have the misclassification rate of positive samples as a prior knowledge. We demonstrate the effectiveness of the proposed method through a synthetic experiment with simple linear models and benchmark problems with neural network models. We also apply our method to drivers’ drowsiness prediction to show that it works well with a real-world problem where confidence is obtained based on manual annotation.

Graph Neural Network (GNN) is a powerful model to learn representations and make predictions on graph data. Existing efforts on GNN have largely defined the graph convolution as a weighted sum of the features of the connected nodes to form the representation of the target node. Nevertheless, the operation of weighted sum assumes the neighbor nodes are independent of each other, and ignores the possible interactions between them. When such interactions exist, such as the co-occurrence of two neighbor nodes is a strong signal of the target node's characteristics, existing GNN models may fail to capture the signal. In this work, we argue the importance of modeling the interactions between neighbor nodes in GNN. We propose a new graph convolution operator, which augments the weighted sum with pairwise interactions of the representations of neighbor nodes. We term this framework as Bilinear Graph Neural Network (BGNN), which improves GNN representation ability with bilinear interactions between neighbor nodes. In particular, we specify two BGNN models named BGCN and BGAT, based on the well-known GCN and GAT, respectively. Empirical results on three public benchmarks of semi-supervised node classification verify the effectiveness of BGNN --- BGCN (BGAT) outperforms GCN (GAT) by 1.6% (1.5%) in classification accuracy. Codes are available at: https://github.com/zhuhm1996/bgnn.

Recent work on bidirectional search defined a lower bound on costs of paths between pairs of nodes, and introduced a new algorithm, NBS, which is based on this bound. Building on these results, we introduce DVCBS, a new algorithm that aims to to further reduce the number of expansions. Generalizing beyond specific algorithms, we then propose a method for enhancing heuristics by propagating such lower bounds (lb-propagation) between frontiers. This lb-propagation can be used in existing algorithms, often improving their performance, as well as making them "well behaved".

Semi-supervised domain adaptation (SSDA) is a novel branch of machine learning that scarce labeled target examples are available, compared with unsupervised domain adaptation. To make effective use of these additional data so as to bridge the domain gap, one possible way is to generate adversarial examples, which are images with additional perturbations, between the two domains and fill the domain gap. Adversarial training has been proven to be a powerful method for this purpose. However, the traditional adversarial training adds noises in arbitrary directions, which is inefficient to migrate between domains, or generate directional noises from the source to target domain and reverse. In this work, we devise a general bidirectional adversarial training method and employ gradient to guide adversarial examples across the domain gap, i.e., the Adaptive Adversarial Training (AAT) for source to target domain and Entropy-penalized Virtual Adversarial Training (E-VAT) for target to source domain. Particularly, we devise a Bidirectional Adversarial Training (BiAT) network to perform diverse adversarial trainings jointly. We evaluate the effectiveness of BiAT on three benchmark datasets and experimental results demonstrate the proposed method achieves the state-of-the-art.

We investigate opinion dynamics in multi-agent networks when there exists a bias toward one of two possible opinions; for example, reflecting a status quo vs a superior alternative. Starting with all agents sharing an initial opinion representing the status quo, the system evolves in steps. In each step, one agent selected uniformly at random adopts with some probability a the superior opinion, and with probability 1 - a it follows an underlying update rule to revise its opinion on the basis of those held by its neighbors. We analyze the convergence of the resulting process under two well-known update rules, namely majority and voter. The framework we propose exhibits a rich structure, with a nonobvious interplay between topology and underlying update rule. For example, for the voter rule we show that the speed of convergence bears no significant dependence on the underlying topology, whereas the picture changes completely under the majority rule, where network density negatively affects convergence. We believe that the model we propose is at the same time simple, rich, and modular, affording mathematical characterization of the interplay between bias, underlying opinion dynamics, and social structure in a unified setting.

To address the challenges posed by unknown occlusions, we propose a Biased Feature Learning (BFL) framework for occlusion-invariant face recognition. We first construct an extended dataset using a multi-scale data augmentation method. For model training, we modify the label loss to adjust the impact of normal and occluded samples. Further, we propose a biased guidance strategy to manipulate the optimization of a network so that the feature embedding space is dominated by non-occluded faces. BFL not only enhances the robustness of a network to unknown occlusions but also maintains or even improves its performance for normal faces. Experimental results demonstrate its superiority as well as the generalization capability with different network architectures and loss functions.

We address the challenging issue of deformable registration that robustly and efficiently builds dense correspondences between images. Traditional approaches upon iterative energy optimization typically invoke expensive computational load. Recent learning-based methods are able to efficiently predict deformation maps by incorporating learnable deep networks. Unfortunately, these deep networks are designated to learn deterministic features for classification tasks, which are not necessarily optimal for registration. In this paper, we propose a novel bi-level optimization model that enables jointly learning deformation maps and features for image registration. The bi-level model takes the energy for deformation computation as the upper-level optimization while formulates the maximum \emph{a posterior} (MAP) for features as the lower-level optimization. Further, we design learnable deep networks to simultaneously optimize the cooperative bi-level model, yielding robust and efficient registration. These deep networks derived from our bi-level optimization constitute an unsupervised end-to-end framework for learning both features and deformations. Extensive experiments of image-to-atlas and image-to-image deformable registration on 3D brain MR datasets demonstrate that we achieve state-of-the-art performance in terms of accuracy, efficiency, and robustness.

Network pruning has been proposed as a remedy for alleviating the over-parameterization problem of deep neural networks. However, its value has been recently challenged especially from the perspective of neural architecture search (NAS). We challenge the conventional wisdom of pruning-after-training by proposing a joint search-and-training approach that directly learns a compact network from the scratch. By treating pruning as a search strategy, we present two new insights in this paper: 1) it is possible to expand the search space of networking pruning by associating each filter with a learnable weight; 2) joint search-and-training can be conducted iteratively to maximize the learning efficiency. More specifically, we propose a coarse-to-fine tuning strategy to iteratively sample and update compact sub-network to approximate the target network. The weights associated with network filters will be accordingly updated by joint search-and-training to reflect learned knowledge in NAS space. Moreover, we introduce strategies of random perturbation (inspired by Monte Carlo) and flexible thresholding (inspired by Reinforcement Learning) to adjust the weight and size of each layer. Extensive experiments on ResNet and VGGNet demonstrate the superior performance of our proposed method on popular datasets including CIFAR10, CIFAR100 and ImageNet.

Knowledge present in a domain is well expressed as relationships between corresponding concepts. For example, in zoology, animal species form complex hierarchies; in genomics, the different (parts of) molecules are organized in groups and subgroups based on their functions; plants, molecules, and astronomical objects all form complex taxonomies. Nevertheless, when applying supervised machine learning (ML) in such domains, we commonly reduce the complex and rich knowledge to a fixed set of labels. This oversimplifies and limits the potential impact that the ML solution can deliver. The main reason for such a reductionist approach is the difficulty in eliciting the domain knowledge from the experts. Developing a label structure with sufficient fidelity and providing comprehensive multi-label annotation can be exceedingly labor-intensive in many real-world applications. Here, we provide a method for efficient hierarchical knowledge elicitation (HKE) from experts working with high-dimensional data such as images or videos. Our method is based on psychometric testing and active deep metric learning. The developed models embed the high-dimensional data in a metric space where distances are semantically meaningful, and the data can be organized in a hierarchical structure.

An efficient and effective person re-identification (ReID) system relieves the users from painful and boring video watching and accelerates the process of video analysis. Recently, with the explosive demands of practical applications, a lot of research efforts have been dedicated to heterogeneous person re-identification (Hetero-ReID). In this paper, we provide a comprehensive review of state-of-the-art Hetero-ReID methods that address the challenge of inter-modality discrepancies. According to the application scenario, we classify the methods into four categories --- low-resolution, infrared, sketch, and text. We begin with an introduction of ReID, and make a comparison between Homogeneous ReID (Homo-ReID) and Hetero-ReID tasks. Then, we describe and compare existing datasets for performing evaluations, and survey the models that have been widely employed in Hetero-ReID. We also summarize and compare the representative approaches from two perspectives, i.e., the application scenario and the learning pipeline. We conclude by a discussion of some future research directions. Follow-up updates are available at https://github.com/lightChaserX/Awesome-Hetero-reID

AMR-to-text generation is a challenging task of generating texts from graph-based semantic representations. Recent studies formalize this task a graph-to-sequence learning problem and use various graph neural networks to model graph structure. In this paper, we propose a novel approach that generates texts from AMR graphs while reconstructing the input graph structures. Our model employs graph attention mechanism to aggregate information for encoding the inputs. Moreover, better node representations are learned by optimizing two simple but effective auxiliary reconstruction objectives: link prediction objective which requires predicting the semantic relationship between nodes, and distance prediction objective which requires predicting the distance between nodes. Experimental results on two benchmark datasets show that our proposed model improves considerably over strong baselines and achieves new state-of-the-art.

State estimation methods based on hybrid discrete and continuous state models have emerged as a method of precisely computing belief states for real world systems, however they have difficulty scaling to systems with more than a handful of components. Classical, consistency based diagnosis methods scale to this level by combining best-first enumeration and conflict-directed search. While best-first methods have been developed for hybrid estimation, conflict-directed methods have thus far been elusive as conflicts summarize constraint violations, but probabilistic hybrid estimation is relatively unconstrained. In this paper we present an approach (A*BC) that unifies best-first enumeration and conflict-directed search in relatively unconstrained problems through the concept of "bounding" conflicts, an extension of conflicts that represent tighter bounds on the cost of regions of the search space. Experiments show that an A*BC powered state estimator produces estimates up to an order of magnitude faster than the current state of the art, particularly on large systems.

We study best-arm identification with fixed confidence in bandit models with graph smoothness constraint. We provide and analyze an efficient gradient ascent algorithm to compute the sample complexity of this problem as a solution of a non-smooth max-min problem (providing in passing a simplified analysis for the unconstrained case). Building on this algorithm, we propose an asymptotically optimal strategy. We furthermore illustrate by numerical experiments both the strategy's efficiency and the impact of the smoothness constraint on the sample complexity. Best Arm Identification (BAI) is an important challenge in many applications ranging from parameter tuning to clinical trials. It is now very well understood in vanilla bandit models, but real-world problems typically involve some dependency between arms that requires more involved models. Assuming a graph structure on the arms is an elegant practical way to encompass this phenomenon, but this had been done so far only for regret minimization. Addressing BAI with graph constraints involves delicate optimization problems for which the present paper offers a solution.

We study how belief merging operators can be considered as maximum likelihood estimators, i.e., we assume that there exists a (unknown) true state of the world and that each agent participating in the merging process receives a noisy signal of it, characterized by a noise model. The objective is then to aggregate the agents' belief bases to make the best possible guess about the true state of the world. In this paper, some logical connections between the rationality postulates for belief merging (IC postulates) and simple conditions over the noise model under consideration are exhibited. These results provide a new justification for IC merging postulates. We also provide results for two specific natural noise models: the world swap noise and the atom swap noise, by identifying distance-based merging operators that are maximum likelihood estimators for these two noise models.

Bayesian optimization (BO) is a popular approach for expensive black-box optimization, with applications including parameter tuning, experimental design, and robotics. BO usually models the objective function by a Gaussian process (GP), and iteratively samples the next data point by maximizing an acquisition function. In this paper, we propose a new general framework for BO by generating pseudo-points (i.e., data points whose objective values are not evaluated) to improve the GP model. With the classic acquisition function, i.e., upper confidence bound (UCB), we prove that the cumulative regret can be generally upper bounded. Experiments using UCB and other acquisition functions, i.e., probability of improvement (PI) and expectation of improvement (EI), on synthetic as well as real-world problems clearly show the advantage of generating pseudo-points.

The performance of clustering depends on an appropriately defined similarity between two items. When the similarity is measured based on human perception, human workers are often employed to estimate a similarity score between items in order to support clustering, leading to a procedure called crowdsourced clustering. Assuming a monetary reward is paid to a worker for each similarity score and assuming the similarities between pairs and workers' reliability have a large diversity, when the budget is limited, it is critical to wisely assign pairs of items to different workers to optimize the clustering result. We model this budget allocation problem as a Markov decision process where item pairs are dynamically assigned to workers based on the historical similarity scores they provided. We propose an optimistic knowledge gradient policy where the assignment of items in each stage is based on the minimum-weight K-cut defined on a similarity graph. We provide simulation studies and real data analysis to demonstrate the performance of the proposed method.

Smart agriculture (SmartAg) has emerged as a rich domain for AI-driven decision support systems (DSS); however, it is often challenged by user-adoption issues. This paper reports a case-based reasoning (CBR) system, PBI-CBR, that predicts grass growth for dairy farmers, that combines predictive accuracy and explanations to improve user adoption. PBI-CBR’s key novelty is its use of Bayesian methods for case-base maintenance in a regression domain. Experiments report the tradeoff between predictive accuracy and explanatory capability for different variants of PBI-CBR, and how updating Bayesian priors each year improves performance.