arXiv subjects

Hong Chen

Publications and source records attributed to Hong Chen.

At least 217 records · Page 12Linked to original sources

arXiv2022

Subgraph Retrieval Enhanced Model for Multi-hop Knowledge Base Question Answering

Recent works on knowledge base question answering (KBQA) retrieve subgraphs for easier reasoning. A desired subgraph is crucial as a small one may exclude the answer but a large one might introduce more noises. However, the existing retrieval is either heuristic or interwoven with the reasoning, causing reasoning on the partial subgraphs, which increases the reasoning bias when the intermediate supervision is missing. This paper proposes a trainable subgraph retriever (SR) decoupled from the subsequent reasoning process, which enables a plug-and-play framework to enhance any subgraph-oriented KBQA model. Extensive experiments demonstrate SR achieves significantly better retrieval and QA performance than existing retrieval methods. Via weakly supervised pre-training as well as the end-to-end fine-tuning, SRl achieves new state-of-the-art performance when combined with NSM, a subgraph-oriented reasoner, for embedding-based KBQA methods.

cs.CL↗

arXiv2022

Self-supervised Graph Learning for Occasional Group Recommendation

As an important branch in Recommender System, occasional group recommendation has received more and more attention. In this scenario, each occasional group (cold-start group) has no or few historical interacted items. As each occasional group has extremely sparse interactions with items, traditional group recommendation methods can not learn high-quality group representations. The recent proposed Graph Neural Networks (GNNs), which incorporate the high-order neighbors of the target occasional group, can alleviate the above problem in some extent. However, these GNNs still can not explicitly strengthen the embedding quality of the high-order neighbors with few interactions. Motivated by the Self-supervised Learning technique, which is able to find the correlations within the data itself, we propose a self-supervised graph learning framework, which takes the user/item/group embedding reconstruction as the pretext task to enhance the embeddings of the cold-start users/items/groups. In order to explicitly enhance the high-order cold-start neighbors' embedding quality, we further introduce an embedding enhancer, which leverages the self-attention mechanism to improve the embedding quality for them. Comprehensive experiments show the advantages of our proposed framework than the state-of-the-art methods.

cs.IR↗

arXiv2022

No One Left Behind: Inclusive Federated Learning over Heterogeneous Devices

Federated learning (FL) is an important paradigm for training global models from decentralized data in a privacy-preserving way. Existing FL methods usually assume the global model can be trained on any participating client. However, in real applications, the devices of clients are usually heterogeneous, and have different computing power. Although big models like BERT have achieved huge success in AI, it is difficult to apply them to heterogeneous FL with weak clients. The straightforward solutions like removing the weak clients or using a small model to fit all clients will lead to some problems, such as under-representation of dropped clients and inferior accuracy due to data loss or limited model representation ability. In this work, we propose InclusiveFL, a client-inclusive federated learning method to handle this problem. The core idea of InclusiveFL is to assign models of different sizes to clients with different computing capabilities, bigger models for powerful clients and smaller ones for weak clients. We also propose an effective method to share the knowledge among multiple local models with different sizes. In this way, all the clients can participate in the model learning in FL, and the final model can be big and powerful enough. Besides, we propose a momentum knowledge distillation method to better transfer knowledge in big models on powerful clients to the small models on weak clients. Extensive experiments on many real-world benchmark datasets demonstrate the effectiveness of the proposed method in learning accurate models from clients with heterogeneous devices under the FL framework.

cs.LG↗

arXiv2022

Lessons learned from the NeurIPS 2021 MetaDL challenge: Backbone fine-tuning without episodic meta-learning dominates for few-shot learning image classification

Although deep neural networks are capable of achieving performance superior to humans on various tasks, they are notorious for requiring large amounts of data and computing resources, restricting their success to domains where such resources are available. Metalearning methods can address this problem by transferring knowledge from related tasks, thus reducing the amount of data and computing resources needed to learn new tasks. We organize the MetaDL competition series, which provide opportunities for research groups all over the world to create and experimentally assess new meta-(deep)learning solutions for real problems. In this paper, authored collaboratively between the competition organizers and the top-ranked participants, we describe the design of the competition, the datasets, the best experimental results, as well as the top-ranked methods in the NeurIPS 2021 challenge, which attracted 15 active teams who made it to the final phase (by outperforming the baseline), making over 100 code submissions during the feedback phase. The solutions of the top participants have been open-sourced. The lessons learned include that learning good representations is essential for effective transfer learning.

cs.LG↗

arXiv2022

The existence of multi-peak positive solutions for nonlinear Kirchhoff equations

In this work, we study the following Kirchhoff equation $$\begin{cases}-\left(\varepsilon^2 a+\varepsilon b\int_{\mathbb R^3}|\nabla u|^2\right)Δu +u =Q(x)u^{q-1},\quad u>0,\quad x\in {\mathbb{R}^{3}},\\u\to 0,\quad \text{as}\ |x|\to +\infty,\end{cases}$$ where $a,b>0$ are constants, $2 0$ is a parameter. Under some suitable assumptions on the function $Q(x)$, we obtain that the equation above has positive multi-peak solutions concentrating at a critical point of $Q(x)$ for $\varepsilon>0$ sufficiently small, by using the Lyapunov-Schmidt reduction method. We extend the result in (Discrete Contin. Dynam. Systems 6(2000), 39--50) to the nonlinear Kirchhoff equation.

math.AP↗

arXiv2022

How zealots affect the energy cost for controlling complex social networks

The controllability of complex networks may be applicable for understanding how to control a complex social network, where members share their opinions and influence one another. Previous works in this area have focused on controllability, energy cost, or optimization under the assumption that all nodes are compliant, passing on information neutrally without any preferences. However, the assumption on nodal neutrality should be reassessed, given that in networked social systems, some people may hold fast to their personal beliefs. By introducing some stubborn agents, or zealots, who hold steadfast to their beliefs and seek to influence others, the control energy is computed and compared against those without zealots. It was found that the presence of zealots alters the energy cost at a quadratic rate with respect to their own fixed beliefs. However, whether or not the zealots' presence increases or decreases the energy cost is affected by the interplay between different parameters such as the zealots' beliefs, number of drivers, final control time regimes, network effects, network dynamics, and number and configurations of neutral nodes influenced by the zealots. For example, when a network dynamics is linear but does not have conformity behavior, it could be possible for a contrarian zealot to assist in reducing control energy. With conformity behavior, a contrarian zealot always negatively affects network control by increasing energy cost. The results of this paper suggest caution when modeling real networked social systems with the controllability of networked linear dynamics since the system dynamical behavior is sensitive to parameter change.

physics.soc-ph↗

arXiv2022

A Multi-Strategy based Pre-Training Method for Cold-Start Recommendation

Cold-start problem is a fundamental challenge for recommendation tasks. The recent self-supervised learning (SSL) on Graph Neural Networks (GNNs) model, PT-GNN, pre-trains the GNN model to reconstruct the cold-start embeddings and has shown great potential for cold-start recommendation. However, due to the over-smoothing problem, PT-GNN can only capture up to 3-order relation, which can not provide much useful auxiliary information to depict the target cold-start user or item. Besides, the embedding reconstruction task only considers the intra-correlations within the subgraph of users and items, while ignoring the inter-correlations across different subgraphs. To solve the above challenges, we propose a multi-strategy based pre-training method for cold-start recommendation (MPT), which extends PT-GNN from the perspective of model architecture and pretext tasks to improve the cold-start recommendation performance. Specifically, in terms of the model architecture, in addition to the short-range dependencies of users and items captured by the GNN encoder, we introduce a Transformer encoder to capture long-range dependencies. In terms of the pretext task, in addition to considering the intra-correlations of users and items by the embedding reconstruction task, we add embedding contrastive learning task to capture inter-correlations of users and items. We train the GNN and Transformer encoders on these pretext tasks under the meta-learning setting to simulate the real cold-start scenario, making the model easily and rapidly being adapted to new cold-start users and items. Experiments on three public recommendation datasets show the superiority of the proposed MPT model against the vanilla GNN models, the pre-training GNN model on user/item embedding inference and the recommendation task.

cs.IR↗

arXiv2022

Injecting Numerical Reasoning Skills into Knowledge Base Question Answering Models

Embedding-based methods are popular for Knowledge Base Question Answering (KBQA), but few current models have numerical reasoning skills and thus struggle to answer ordinal constrained questions. This paper proposes a new embedding-based KBQA framework which particularly takes numerical reasoning into account. We present NumericalTransformer on top of NSM, a state-of-the-art embedding-based KBQA model, to create NT-NSM. To enable better training, we propose two pre-training tasks with explicit numerical-oriented loss functions on two generated training datasets and a template-based data augmentation method for enriching ordinal constrained QA dataset. Extensive experiments on KBQA benchmarks demonstrate that with the help of our training algorithm, NT-NSM is empowered with numerical reasoning skills and substantially outperforms the baselines in answering ordinal constrained questions.

cs.CL↗

arXiv2022

Photonic Dirac Nodal Line Semimetals Realized by Hyper-crystal

Recently, the gapless Dirac/Weyl nodal semimetals with linear dispersion and topologically protected modes degeneracy are rapidly growing frontiers of topological physics. Especially, type-I, type-II, and critical type-III nodal semimetals are discovered according to the tilt angles of the Dirac/Weyl cones. Here, by introducing hyperbolic metamaterials into one-dimensional photonic crystals, we design the "hyper-crystal" and study the photonic four-fold degenerate Dirac nodal line semimetals (DNLSs) with two types of perpendicularly polarized waves. Moreover, the flexibly controlled photonic DNLSs using the phase compensation effect of hyperbolic dispersion are studied. Our results not only demonstrate a new platform to realize the various photonic DNLSs, where the optical polarization plays the role of electron spin in electronic DNLSs, but also may pave a novel way to explore the abundant Dirac/Weyl physics in the simple classical wave systems.

physics.optics↗

arXiv2022

OMG: Observe Multiple Granularities for Natural Language-Based Vehicle Retrieval

Retrieving tracked-vehicles by natural language descriptions plays a critical role in smart city construction. It aims to find the best match for the given texts from a set of tracked vehicles in surveillance videos. Existing works generally solve it by a dual-stream framework, which consists of a text encoder, a visual encoder and a cross-modal loss function. Although some progress has been made, they failed to fully exploit the information at various levels of granularity. To tackle this issue, we propose a novel framework for the natural language-based vehicle retrieval task, OMG, which Observes Multiple Granularities with respect to visual representation, textual representation and objective functions. For the visual representation, target features, context features and motion features are encoded separately. For the textual representation, one global embedding, three local embeddings and a color-type prompt embedding are extracted to represent various granularities of semantic features. Finally, the overall framework is optimized by a cross-modal multi-granularity contrastive loss function. Experiments demonstrate the effectiveness of our method. Our OMG significantly outperforms all previous methods and ranks the 9th on the 6th AI City Challenge Track2. The codes are available at https://github.com/dyhBUPT/OMG.

cs.CV↗

arXiv2022

Go Back in Time: Generating Flashbacks in Stories with Event Temporal Prompts

Stories or narratives are comprised of a sequence of events. To compose interesting stories, professional writers often leverage a creative writing technique called flashback that inserts past events into current storylines as we commonly observe in novels and plays. However, it is challenging for machines to generate flashback as it requires a solid understanding of event temporal order (e.g. "feeling hungry" before "eat," not vice versa), and the creativity to arrange storylines so that earlier events do not always appear first in narrative order. Two major issues in existing systems that exacerbate the challenges: 1) temporal bias in pertaining and story datasets that leads to monotonic event temporal orders; 2) lack of explicit guidance that helps machines decide where to insert flashbacks. We propose to address these issues using structured storylines to encode events and their pair-wise temporal relations (before, after and vague) as temporal prompts that guide how stories should unfold temporally. We leverage a Plan-and-Write framework enhanced by reinforcement learning to generate storylines and stories end-to-end. Evaluation results show that the proposed method can generate more interesting stories with flashbacks while maintaining textual diversity, fluency, and temporal coherence.

cs.CL↗

arXiv2022

NOC-REK: Novel Object Captioning with Retrieved Vocabulary from External Knowledge

Novel object captioning aims at describing objects absent from training data, with the key ingredient being the provision of object vocabulary to the model. Although existing methods heavily rely on an object detection model, we view the detection step as vocabulary retrieval from an external knowledge in the form of embeddings for any object's definition from Wiktionary, where we use in the retrieval image region features learned from a transformers model. We propose an end-to-end Novel Object Captioning with Retrieved vocabulary from External Knowledge method (NOC-REK), which simultaneously learns vocabulary retrieval and caption generation, successfully describing novel objects outside of the training dataset. Furthermore, our model eliminates the requirement for model retraining by simply updating the external knowledge whenever a novel object appears. Our comprehensive experiments on held-out COCO and Nocaps datasets show that our NOC-REK is considerably effective against SOTAs.

cs.CV↗

arXiv2022

FLDP: Flexible strategy for local differential privacy

Local differential privacy (LDP), a technique applying unbiased statistical estimations instead of real data, is often adopted in data collection. In particular, this technique is used with frequency oracles (FO) because it can protect each user's privacy and prevent leakage of sensitive information. However, the definition of LDP is so conservative that it requires all inputs to be indistinguishable after perturbation. Indeed, LDP protects each value; however, it is rarely used in practical scenarios owing to its cost in terms of accuracy. In this paper, we address the challenge of providing weakened but flexible protection where each value only needs to be indistinguishable from part of the domain after perturbation. First, we present this weakened but flexible LDP (FLDP) notion. We then prove the association with LDP and DP. Second, we design an FHR approach for the common FO issue while satisfying FLDP. The proposed approach balances communication cost, computational complexity, and estimation accuracy. Finally, experimental results using practical and synthetic datasets verify the effectiveness and efficiency of our approach.

cs.CR↗

arXiv2022

Regularized Modal Regression on Markov-dependent Observations: A Theoretical Assessment

Modal regression, a widely used regression protocol, has been extensively investigated in statistical and machine learning communities due to its robustness to outliers and heavy-tailed noises. Understanding modal regression's theoretical behavior can be fundamental in learning theory. Despite significant progress in characterizing its statistical property, the majority of the results are based on the assumption that samples are independent and identical distributed (i.i.d.), which is too restrictive for real-world applications. This paper concerns the statistical property of regularized modal regression (RMR) within an important dependence structure - Markov dependent. Specifically, we establish the upper bound for RMR estimator under moderate conditions and give an explicit learning rate. Our results show that the Markov dependence impacts on the generalization error in the way that sample size would be discounted by a multiplicative factor depending on the spectral gap of underlying Markov chain. This result shed a new light on characterizing the theoretical underpinning for robust regression.

stat.ML↗

arXiv2022

Error-based Knockoffs Inference for Controlled Feature Selection

Recently, the scheme of model-X knockoffs was proposed as a promising solution to address controlled feature selection under high-dimensional finite-sample settings. However, the procedure of model-X knockoffs depends heavily on the coefficient-based feature importance and only concerns the control of false discovery rate (FDR). To further improve its adaptivity and flexibility, in this paper, we propose an error-based knockoff inference method by integrating the knockoff features, the error-based feature importance statistics, and the stepdown procedure together. The proposed inference procedure does not require specifying a regression model and can handle feature selection with theoretical guarantees on controlling false discovery proportion (FDP), FDR, or k-familywise error rate (k-FWER). Empirical evaluations demonstrate the competitive performance of our approach on both simulated and real data.

stat.ME↗

arXiv2022

Markov subsampling based Huber Criterion

Subsampling is an important technique to tackle the computational challenges brought by big data. Many subsampling procedures fall within the framework of importance sampling, which assigns high sampling probabilities to the samples appearing to have big impacts. When the noise level is high, those sampling procedures tend to pick many outliers and thus often do not perform satisfactorily in practice. To tackle this issue, we design a new Markov subsampling strategy based on Huber criterion (HMS) to construct an informative subset from the noisy full data; the constructed subset then serves as a refined working data for efficient processing. HMS is built upon a Metropolis-Hasting procedure, where the inclusion probability of each sampling unit is determined using the Huber criterion to prevent over scoring the outliers. Under mild conditions, we show that the estimator based on the subsamples selected by HMS is statistically consistent with a sub-Gaussian deviation bound. The promising performance of HMS is demonstrated by extensive studies on large scale simulations and real data examples.

stat.ML↗

arXiv2022

Equilibrium Defaultable Corporate Debt and Investment

In dynamic capital structure models with an investor break-even condition, the firm's Bellman equation may not generate a contraction mapping, so the standard existence and uniqueness conditions do not apply. First, we provide an example showing the problem in a classical trade-off model. The firm can issue one-period defaultable debt, invest in capital and pay a dividend. If the firm cannot meet the required debt payment, it is liquidated. Second, we show how to use a dual to the original problem and a change of measure, such that existence and uniqueness can be proved. In the unique Markov-perfect equilibrium, firm decisions reflect state-dependent capital and debt targets. Our approach may be useful for other dynamic firm models that have an investor break-even condition.

q-fin.GN↗

arXiv2022

Local well-posedness to the 2D Cauchy problem of full compressible magnetohydrodynamic equations with vacuum at infinity

This paper concerns the Cauchy problem of two-dimensional (2D) full compressible magnetohydrodynamic (MHD) equations in the whole plane $\mathbb{R}^2$ with zero density at infinity. By spatial weighted energy method, we derive the local existence and uniqueness of strong solutions provided that the initial density and the initial magnetic field decay not too slowly at infinity. Note that the initial temperature does not need to decay slowly at infinity. In particular, vacuum states at both the interior domain and the far field are allowed.

math.AP↗