Yuhong Guo

AAAI Conference 2025 Conference Paper

Bi-Level Optimization for Semi-Supervised Learning with Pseudo-Labeling

• Marzi Heidari
• Yuhong Guo

Semi-supervised learning (SSL) is a fundamental task in machine learning, empowering models to extract valuable insights from datasets with limited labeled samples and a large amount of unlabeled data. Although pseudo-labeling is a widely used approach for SSL that generates pseudo-labels for unlabeled data and leverages them as ground truth labels for training, traditional pseudo-labeling techniques often face challenges that significantly decrease the quality of pseudo-labels and hence the overall model performance. In this paper, we propose a novel Bi-level Optimization method for Pseudo-label Learning (BOPL) to boost semi-supervised training. It treats pseudo-labels as latent variables, and optimizes the model parameters and pseudo-labels jointly within a bi-level optimization framework. By enabling direct optimization over the pseudo-labels towards maximizing the prediction model performance, the method is expected to produce high-quality pseudo-labels. To evaluate the effectiveness of the proposed approach, we conduct extensive experiments on multiple SSL benchmarks. The experimental results show the proposed BOPL outperforms the state-of-the-art SSL techniques.

NeurIPS Conference 2025 Conference Paper

Learning to Clean: Reinforcement Learning for Noisy Label Correction

• Marzi Heidari
• Hanping Zhang
• Yuhong Guo

The challenge of learning with noisy labels is significant in machine learning, as it can severely degrade the performance of prediction models if not addressed properly. This paper introduces a novel framework that conceptualizes noisy label correction as a reinforcement learning (RL) problem. The proposed approach, Reinforcement Learning for Noisy Label Correction (RLNLC), defines a comprehensive state space representing data and their associated labels, an action space that indicates possible label corrections, and a reward mechanism that evaluates the efficacy of label corrections. RLNLC learns a deep feature representation based policy network to perform label correction through reinforcement learning, utilizing an actor-critic method. The learned policy is subsequently deployed to iteratively correct noisy training labels and facilitate the training of the prediction model. The effectiveness of RLNLC is demonstrated through extensive experiments on multiple benchmark datasets, where it consistently outperforms existing state-of-the-art techniques for learning with noisy labels.

ECAI Conference 2025 Conference Paper

Skill-Enhanced Reinforcement Learning Acceleration from Heterogeneous Demonstrations

• Hanping Zhang
• Yuhong Guo

Learning from Demonstration (LfD) is a well-established problem in Reinforcement Learning (RL), which aims to facilitate rapid RL by leveraging expert demonstrations to pre-train the RL agent. However, the limited availability of expert demonstration data often hinders its ability to effectively aid downstream RL learning. To address this problem, we propose a novel two-stage method dubbed as Skill-enhanced Reinforcement Learning Acceleration (SeRLA). SeRLA introduces a skill-level adversarial Positive-Unlabeled (PU) learning model that extracts useful skill prior knowledge by learning from both expert demonstrations and general low-cost demonstrations in the offline prior learning stage. Building on this, it employs a skill-based soft actor-critic algorithm to leverage the acquired priors for efficient training of a skill policy network in the downstream online RL stage. In addition, we propose a simple skill-level data enhancement technique to mitigate data sparsity and further improve both skill prior learning and skill policy training. Experiments across multiple standard RL benchmarks demonstrate that SeRLA achieves state-of-the-art performance in accelerating reinforcement learning on downstream tasks, particularly in the early training phase.

AAAI Conference 2024 Conference Paper

Federated Partial Label Learning with Local-Adaptive Augmentation and Regularization

• Yan Yan
• Yuhong Guo

Partial label learning (PLL) expands the applicability of supervised machine learning models by enabling effective learning from weakly annotated overcomplete labels. Existing PLL methods however focus on the standard centralized learning scenarios. In this paper, we expand PLL into the distributed computation setting by formalizing a new learning scenario named as federated partial label learning (FedPLL), where the training data with partial labels are distributed across multiple local clients with privacy constraints. To address this challenging problem, we propose a novel Federated PLL method with Local-Adaptive Augmentation and Regularization (FedPLL-LAAR). In addition to alleviating the partial label noise with moving-average label disambiguation, the proposed method performs MixUp-based local-adaptive data augmentation to mitigate the challenge posed by insufficient and imprecisely annotated local data, and dynamically incorporates the guidance of global model to minimize client drift through adaptive gradient alignment regularization between the global and local models. Extensive experiments conducted on multiple datasets under the FedPLL setting demonstrate the effectiveness of the proposed FedPLL-LAAR method for federated partial label learning.

NeurIPS Conference 2024 Conference Paper

Reinforcement Learning Guided Semi-Supervised Learning

• Marzi Heidari
• Hanping Zhang
• Yuhong Guo

In recent years, semi-supervised learning (SSL) has gained significant attention due to its ability to leverage both labeled and unlabeled data to improve model performance, especially when labeled data is scarce. However, most current SSL methods rely on heuristics or predefined rules for generating pseudo-labels and leveraging unlabeled data. They are limited to exploiting loss functions and regularization methods within the standard norm. In this paper, we propose a novel Reinforcement Learning (RL) Guided SSL method, RLGSSL, that formulates SSL as a one-armed bandit problem and deploys an innovative RL loss based on weighted reward to adaptively guide the learning process of the prediction model. RLGSSL incorporates a carefully designed reward function that balances the use of labeled and unlabeled data to enhance generalization performance. A semi-supervised teacher-student framework is further deployed to increase the learning stability. We demonstrate the effectiveness of RLGSSL through extensive experiments on several benchmark datasets and show that our approach achieves consistent superior performance compared to state-of-the-art SSL methods.

ECAI Conference 2023 Conference Paper

Evolving Dictionary Representation for Few-Shot Class-Incremental Learning

• Xuejun Han
• Yuhong Guo

New objects are continuously emerging in the dynamically changing world and a real-world artificial intelligence system should be capable of continual and effectual adaptation to new emerging classes without forgetting old ones. In view of this, in this paper we tackle a challenging and practical continual learning scenario named few-shot class-incremental learning (FSCIL), in which labeled data are given for classes in a base session but very limited labeled instances are available for new incremental classes. To address this problem, we propose a novel and succinct approach by introducing deep dictionary learning which is a hybrid learning architecture that combines dictionary learning and visual representation learning to provide a better space for characterizing different classes. We simultaneously optimize the dictionary and the feature extraction backbone in the base session, but only finetune the dictionary in the incremental session for adaptation to novel classes, which can alleviate the forgetting problem on base classes compared to finetuning the entire model. To further facilitate future adaptation, we also incorporate multiple pseudo classes into the base session training so that certain space projected by dictionary can be reserved for future new concepts. Moreover, we adopt a novel optimization process by implicitly incorporating the dictionary reconstruction loss into the learning process without adding extra loss terms. The extensive experimental results on CIFAR100, miniImageNet and CUB200 validate the effectiveness of our approach compared to other SOTA methods.

ICLR Conference 2023 Conference Paper

Mutual Partial Label Learning with Competitive Label Noise

• Yan Yan 0025
• Yuhong Guo

Partial label learning (PLL) is an important weakly supervised learning problem, where each training instance is associated with a set of candidate labels that include both the true label and additional noisy labels. Most existing PLL methods assume the candidate noisy labels are randomly chosen, which hardly holds in real-world learning scenarios. In this paper, we consider a more realistic PLL scenario with competitive label noise that is more difficult to distinguish from the true label than the random label noise. We propose a novel Mutual Learning based PLL approach named ML-PLL to address this challenging problem. ML-PLL learns a prediction network based classifier and a class-prototype based classifier cooperatively through interactive mutual learning and label correction. Moreover, we use a transformation network to model the association relationships between the true label and candidate labels, and learn it together with the prediction network to match the observed candidate labels in the training data and enhance label correction. Extensive experiments are conducted on several benchmark PLL datasets, and the proposed ML-PLL approach demonstrates state-of-the-art performance for partial label learning.

ICLR Conference 2023 Conference Paper

Partial Label Unsupervised Domain Adaptation with Class-Prototype Alignment

• Yan Yan 0025
• Yuhong Guo

Partial label learning (PLL) tackles the problem where each instance is associated with a set of candidate labels, only one of which is the ground-truth label. Most existing PLL approaches assume that both the training and test sets share an identical data distribution. However, this assumption does not hold in many real-world scenarios where the training and test data come from different distributions. In this paper, we formalize this learning scenario as a new problem called partial label unsupervised domain adaptation (PLUDA). To address this challenging PLUDA problem, we propose a novel Prototype Alignment based PLUDA method named PAPLUDA, which dynamically refines the pseudo-labels of instances from both the source and target domains by consulting the outputs of a teacher-student model in a moving-average manner, and bridges the cross-domain discrepancy through inter-domain class-prototype alignment. In addition, a teacher-student model based contrastive regularization is deployed to enhance prediction stability and hence improve the class-prototypes in both domains for PLUDA. Comprehensive experimental results demonstrate that PAPLUDA achieves state-of-the-art performance on the widely used benchmark datasets.

AAAI Conference 2021 Conference Paper

Adversarial Partial Multi-Label Learning with Label Disambiguation

• Yan Yan
• Yuhong Guo

Partial multi-label learning (PML), which tackles the problem of learning multi-label prediction models from instances with overcomplete noisy annotations, has recently started gaining attention from the research community. In this paper, we propose a novel adversarial learning model, PML-GAN, under a generalized encoder-decoder framework for partial multilabel learning. The PML-GAN model uses a disambiguation network to identify irrelevant labels and uses a multi-label prediction network to map the training instances to their disambiguated label vectors, while deploying a generative adversarial network as an inverse mapping from label vectors to data samples in the input feature space. The learning of the overall model corresponds to a minimax adversarial game, which enhances the correspondence of input features with the output labels in a bi-directional mapping. Extensive experiments are conducted on both synthetic and real-world partial multi-label datasets, while the proposed model demonstrates the state-of-the-art performance.

IJCAI Conference 2021 Conference Paper

Multi-level Generative Models for Partial Label Learning with Non-random Label Noise

• Yan Yan
• Yuhong Guo

Partial label (PL) learning tackles the problem where each training instance is associated with a set of candidate labels that include both the true label and some irrelevant noise labels. In this paper, we propose a novel multi-level generative model for partial label learning (MGPLL), which tackles the PL problem by learning both a label level adversarial generator and a feature level adversarial generator under a bi-directional mapping framework between the label vectors and the data samples. MGPLL uses a conditional noise label generation network to model the non-random noise labels and perform label denoising, and uses a multi-class predictor to map the training instances to the denoised label vectors, while a conditional data feature generator is used to form an inverse mapping from the denoised label vectors to data samples. Both the noise label generator and the data feature generator are learned in an adversarial manner to match the observed candidate labels and data features respectively. We conduct extensive experiments on both synthesized and real-world partial label datasets. The proposed approach demonstrates the state-of-the-art performance for partial label learning.

ICML Conference 2021 Conference Paper

Parameterless Transductive Feature Re-representation for Few-Shot Learning

• Wentao Cui
• Yuhong Guo

Recent literature in few-shot learning (FSL) has shown that transductive methods often outperform their inductive counterparts. However, most transductive solutions, particularly the meta-learning based ones, require inserting trainable parameters on top of some inductive baselines to facilitate transduction. In this paper, we propose a parameterless transductive feature re-representation framework that differs from all existing solutions from the following perspectives. (1) It is widely compatible with existing FSL methods, including meta-learning and fine tuning based models. (2) The framework is simple and introduces no extra training parameters when applied to any architecture. We conduct experiments on three benchmark datasets by applying the framework to both representative meta-learning baselines and state-of-the-art FSL methods. Our framework consistently improves performances in all experiments and refreshes the state-of-the-art FSL results.

YNIMG Journal 2020 Journal Article

Brain status modeling with non-negative projective dictionary learning

• Mingli Zhang
• Christian Desrosiers
• Yuhong Guo
• Budhachandra Khundrakpam
• Noor Al-Sharif
• Greg Kiar
• Pedro Valdes-Sosa
• Jean-Baptiste Poline

Accurate prediction of individuals' brain age is critical to establish a baseline for normal brain development. This study proposes to model brain development with a novel non-negative projective dictionary learning (NPDL) approach, which learns a discriminative representation of multi-modal neuroimaging data for predicting brain age. Our approach encodes the variability of subjects in different age groups using separate dictionaries, projecting features into a low-dimensional manifold such that information is preserved only for the corresponding age group. The proposed framework improves upon previous discriminative dictionary learning methods by incorporating orthogonality and non-negativity constraints, which remove representation redundancy and perform implicit feature selection. We study brain development on multi-modal brain imaging data from the PING dataset (N = 841, age = 3-21 years). The proposed analysis uses our NDPL framework to predict the age of subjects based on cortical measures from T1-weighted MRI and connectome from diffusion weighted imaging (DWI). We also investigate the association between age prediction and cognition, and study the influence of gender on prediction accuracy. Experimental results demonstrate the usefulness of NDPL for modeling brain development.

AAAI Conference 2020 Conference Paper

Dual Adversarial Co-Learning for Multi-Domain Text Classification

• Yuan Wu
• Yuhong Guo

With the advent of deep learning, the performance of text classification models have been improved significantly. Nevertheless, the successful training of a good classification model requires a sufficient amount of labeled data, while it is always expensive and time consuming to annotate data. With the rapid growth of digital data, similar classification tasks can typically occur in multiple domains, while the availability of labeled data can largely vary across domains. Some domains may have abundant labeled data, while in some other domains there may only exist a limited amount (or none) of labeled data. Meanwhile text classification tasks are highly domain-dependent — a text classifier trained in one domain may not perform well in another domain. In order to address these issues, in this paper we propose a novel dual adversarial co-learning approach for multi-domain text classification (MDTC). The approach learns shared-private networks for feature extraction and deploys dual adversarial regularizations to align features across different domains and between labeled and unlabeled data simultaneously under a discrepancy based co-learning framework, aiming to improve the classifiers’ generalization capacity with the learned features. We conduct experiments on multi-domain sentiment classi- fication datasets. The results show the proposed approach achieves the state-of-the-art MDTC performance.

AAAI Conference 2020 Conference Paper

Partial Label Learning with Batch Label Correction

• Yan Yan
• Yuhong Guo

Partial label (PL) learning tackles the problem where each training instance is associated with a set of candidate labels, among which only one is the true label. In this paper, we propose a simple but effective batch-based partial label learning algorithm named PL-BLC, which tackles the partial label learning problem with batch-wise label correction (BLC). PL-BLC dynamically corrects the label confidence matrix of each training batch based on the current prediction network, and adopts a MixUp data augmentation scheme to enhance the underlying true labels against the redundant noisy labels. In addition, it introduces a teacher model through a consistency cost to ensure the stability of the batch-based prediction network update. Extensive experiments are conducted on synthesized and real-world partial label learning datasets, while the proposed approach demonstrates the state-of-the-art performance for partial label learning.

ICML Conference 2020 Conference Paper

Time-aware Large Kernel Convolutions

• Vasileios Lioutas
• Yuhong Guo

To date, most state-of-the-art sequence modeling architectures use attention to build generative models for language based tasks. Some of these models use all the available sequence tokens to generate an attention distribution which results in time complexity of $O(n^2)$. Alternatively, they utilize depthwise convolutions with softmax normalized kernels of size $k$ acting as a limited-window self-attention, resulting in time complexity of $O(k{\cdot}n)$. In this paper, we introduce Time-aware Large Kernel (TaLK) Convolutions, a novel adaptive convolution operation that learns to predict the size of a summation kernel instead of using a fixed-sized kernel matrix. This method yields a time complexity of $O(n)$, effectively making the sequence encoding process linear to the number of tokens. We evaluate the proposed method on large-scale standard machine translation, abstractive summarization and language modeling datasets and show that TaLK Convolutions constitute an efficient improvement over other attention/convolution based approaches.

IJCAI Conference 2018 Conference Paper

Matrix completion with Preference Ranking for Top-N Recommendation

• Zengmao Wang
• Yuhong Guo
• Bo Du

Matrix completion has become a popular method for top-N recommendation due to the low rank nature of sparse rating matrices. However, many existing methods produce top-N recommendations by recovering a user-item matrix solely based on a low rank function or its relaxations, while ignoring other important intrinsic characteristics of the top-N recommendation tasks such as preference ranking over the items. In this paper, we propose a novel matrix completion method that integrates the low rank and preference ranking characteristics of recommendation matrix under a self-recovery model for top-N recommendation. The proposed method is formulated as a joint minimization problem and solved using an ADMM algorithm. We conduct experiments on E-commerce datasets. The experimental results show the proposed approach outperforms several state-of-the-art methods.

AAAI Conference 2018 Conference Paper

Visual Relationship Detection With Deep Structural Ranking

• Kongming Liang
• Yuhong Guo
• Hong Chang
• Xilin Chen

Visual relationship detection aims to describe the interactions between pairs of objects. Different from individual object learning tasks, the number of possible relationships are much larger, which makes it hard to explore only based on the visual appearance of objects. In addition, due to the limited human effort, the annotations for visual relationships are usually incomplete which increases the difficulty of model training and evaluation. In this paper, we propose a novel framework, called Deep Structural Ranking, for visual relationship detection. To complement the representation ability of visual appearance, we integrate multiple cues for predicting the relationships contained in an input image. Moreover, we design a new ranking objective function by enforcing the annotated relationships to have higher relevance scores. Unlike previous works, our proposed method can both facilitate the co-occurrence of relationships and mitigate the incompleteness problem. Experimental results show that our proposed method outperforms the state-of-the-art on the two widely used datasets. We also demonstrate its superiority in detecting zero-shot relationships.

AAAI Conference 2017 Conference Paper

Convex Co-Embedding for Matrix Completion with Predictive Side Information

• Yuhong Guo

Matrix completion as a common problem in many application domains has received increasing attention in the machine learning community. Previous matrix completion methods have mostly focused on exploiting the matrix low-rank property to recover missing entries. Recently, it has been noticed that side information that describes the matrix items can help to improve the matrix completion performance. In this paper, we propose a novel matrix completion approach that exploits side information within a principled co-embedding framework. This framework integrates a low-rank matrix factorization model and a label embedding based prediction model together to derive a convex co-embedding formulation with nuclear norm regularization. We develop a fast proximal gradient descent algorithm to solve this co-embedding problem. The effectiveness of the proposed approach is demonstrated on two types of real world application problems.

IJCAI Conference 2017 Conference Paper

Incomplete Attribute Learning with auxiliary labels

• Kongming Liang
• Yuhong Guo
• Hong Chang
• Xilin Chen

Visual attribute learning is a fundamental and challenging problem for image understanding. Considering the huge semantic space of attributes, it is economically impossible to annotate all their presence or absence for a natural image via crowd-sourcing. In this paper, we tackle the incompleteness nature of visual attributes by introducing auxiliary labels into a novel transductive learning framework. By jointly predicting the attributes from the input images and modeling the relationship of attributes and auxiliary labels, the missing attributes can be recovered effectively. In addition, the proposed model can be solved efficiently in an alternative way by optimizing quadratic programming problems and updating parameters in closed-form solutions. Moreover, we propose and investigate different methods for acquiring auxiliary labels. We conduct experiments on three widely used attribute prediction datasets. The experimental results show that our proposed method can achieve the state-of-the-art performance with access to partially observed attribute annotations.

IJCAI Conference 2017 Conference Paper

Learning Discriminative Recommendation Systems with Side Information

• Feipeng Zhao
• Yuhong Guo

Top-N recommendation systems are useful in many real world applications such as E-commerce platforms. Most previous methods produce top-N recommendations based on the observed user purchase or recommendation activities. Recently, it has been noticed that side information that describes the items can be produced from auxiliary sources and help to improve the performance of top-N recommendation systems; e. g. , side information of the items can be collected from the item reviews. In this paper, we propose a joint discriminative prediction model that exploits both the partially observed user-item recommendation matrix and the item-based side information to build top-N recommendation systems. This joint model aggregates observed user-item recommendation activities to produce the missing user-item recommendation scores while simultaneously training a linear regression model to predict the user-item recommendation scores from auxiliary item features. We evaluate the proposed approach on a number of recommendation datasets. The experimental results show that the proposed joint model is very effective for producing top-N recommendation systems.

IJCAI Conference 2016 Conference Paper

Improving Top-N Recommendation with Heterogeneous Loss

• Feipeng Zhao
• Yuhong Guo

Personalized top-N recommendation systems have great impact on many real world applications such as E-commerce platforms and social networks. Most existing methods produce personalized top-N recommendations by minimizing a specific uniform loss such as pairwise ranking loss or pointwise recovery loss. In this paper, we propose a novel personalized top-N recommendation approach that minimizes a combined heterogeneous loss based on linear self-recovery models. The heterogeneous loss integrates the strengths of both pairwise ranking loss and pointwise recovery loss to provide more informative recommendation predictions. We formulate the learning problem with heterogeneous loss as a constrained convex minimization problem and develop a projected stochastic gradient descent optimization algorithm to solve it. We evaluate the proposed approach on a set of personalized top-N recommendation tasks. The experimental results show the proposed approach outperforms a number of state-of-the-art methods on top-N recommendation.

IJCAI Conference 2016 Conference Paper

Predictive Collaborative Filtering with Side Information

• Feipeng Zhao
• Min Xiao
• Yuhong Guo

Recommender systems have been widely studied in the literature as they have real world impacts in many E-commerce platforms and social networks. Most previous systems are based on the user-item recommendation matrix, which contains users' history recommendation activities on items. In this paper, we propose a novel predictive collaborative filtering approach that exploits both the partially observed user-item recommendation matrix and the item-based side information to produce top-N recommender systems. The proposed approach automatically identifies the most interesting items for each user from his or her non-recommended item pool by aggregating over his or her recommended items via a low-rank coefficient matrix. Moreover, it also simultaneously builds linear regression models from the item-based side information such as item reviews to predict the item recommendation scores for the users. The proposed approach is formulated as a rank constrained joint minimization problem with integrated least squares losses, for which an efficient analytical solution can be derived. To evaluate the proposed learning technique, empirical evaluations on five recommendation tasks are conducted. The experimental results demonstrate the efficacy of the proposed approach comparing to the competing methods.

IJCAI Conference 2015 Conference Paper

Multi-Label Classification with Feature-Aware Non-Linear Label Space Transformation

• Xin Li
• Yuhong Guo

Multi-label classification with many classes has recently drawn a lot of attention. Existing methods address this problem by performing linear label space transformation to reduce the dimension of label space, and then conducting independent regression for each reduced label dimension. These methods however do not capture nonlinear correlations of the multiple labels and may lead to significant information loss in the process of label space reduction. In this paper, we first propose to exploit kernel canonical correlation analysis (KCCA) to capture nonlinear label correlation information and perform nonlinear label space reduction. Then we develop a novel label space reduction method that explicitly combines linear and nonlinear label space transformations based on CCA and KCCA respectively to address multi-label classification with many classes. The proposed method is a feature-aware label transformation method that promotes the label predictability in the transformed label space from the input features. We conduct experiments on a number of multi-label classification datasets. The proposed approach demonstrates good performance, comparing to a number of stateof-the-art label dimension reduction methods.

IJCAI Conference 2015 Conference Paper

Semi-Supervised Multi-Label Learning with Incomplete Labels

• Feipeng Zhao
• Yuhong Guo

The problem of incomplete labels is frequently encountered in many application domains where the training labels are obtained via crowd-sourcing. The label incompleteness significantly increases the difficulty of acquiring accurate multi-label prediction models. In this paper, we propose a novel semi-supervised multi-label method that integrates low-rank label matrix recovery into the manifold regularized vector-valued prediction framework to address multi-label learning with incomplete labels. The proposed method is formulated as a convex but non-smooth joint optimization problem over the latent label matrix and the prediction model parameters. We then develop a fast proximal gradient descent with continuation algorithm to solve it for a global optimal solution. The efficacy of the proposed approach is demonstrated on multiple multi-label datasets, comparing to related methods that handle incomplete labels.

AAAI Conference 2014 Conference Paper

Convex Co-embedding

• Farzaneh Mirzazadeh
• Yuhong Guo
• Dale Schuurmans

We present a general framework for association learning, where entities are embedded in a common latent space to express relatedness via geometry—an approach that underlies the state of the art for link prediction, relation learning, multi-label tagging, relevance retrieval and ranking. Although current approaches rely on local training methods applied to non-convex formulations, we demonstrate how general convex formulations can be achieved for entity embedding, both for standard multi-linear and prototype-distance models. We investigate an efficient optimization strategy that allows scaling. An experimental evaluation reveals the advantages of global training in different case studies.

ICML Conference 2014 Conference Paper

Latent Semantic Representation Learning for Scene Classification

• Xin Li 0013
• Yuhong Guo

The performance of machine learning methods is heavily dependent on the choice of data representation. In real world applications such as scene recognition problems, the widely used low-level input features can fail to explain the high-level semantic label concepts. In this work, we address this problem by proposing a novel patch-based latent variable model to integrate latent contextual representation learning and classification model training in one joint optimization framework. Within this framework, the latent layer of variables bridge the gap between inputs and outputs by providing discriminative explanations for the semantic output labels, while being predictable from the low-level input features. Experiments conducted on standard scene recognition tasks demonstrate the efficacy of the proposed approach, comparing to the state-of-the-art scene recognition methods.

UAI Conference 2014 Conference Paper

Multi-label Image Classification with A Probabilistic Label Enhancement Model

• Xin Li 0013
• Feipeng Zhao
• Yuhong Guo

In this paper, we present a novel probabilistic label enhancement model to tackle multi-label image classification problem. Recognizing multiple objects in images is a challenging problem due to label sparsity, appearance variations of the objects and occlusions. We propose to tackle these difficulties from a novel perspective by constructing auxiliary labels in the output space. Our idea is to exploit label combinations to enrich the label space and improve the label identification capacity in the original label space. In particular, we identify a set of informative label combination pairs by constructing a tree-structured graph in the label space using the maximum spanning tree algorithm, which naturally forms a conditional random field. We then use the produced label pairs as auxiliary new labels to augment the original labels and perform piecewise training under the framework of conditional random fields. In the test phase, max-product message passing is used to perform efficient inference on the tree graph, which integrates the augmented label pair classifiers and the standard individual binary classifiers for multi-label prediction. We evaluate the proposed approach on several image classification datasets. The experimental results demonstrate the superiority of our label enhancement model in terms of both prediction performance and running time comparing to the-stateof-the-art multi-label learning methods.

AAAI Conference 2014 Conference Paper

Semi-Supervised Matrix Completion for Cross-Lingual Text Classification

• Min Xiao
• Yuhong Guo

Cross-lingual text classification is the task of assigning labels to observed documents in a label-scarce target language domain by using a prediction model trained with labeled documents from a label-rich source language domain. Cross-lingual text classification is popularly studied in natural language processing area to reduce the expensive manual annotation effort required in the target language domain. In this work, we propose a novel semi-supervised representation learning approach to address this challenging task by inducing interlingual features via semi-supervised matrix completion. To evaluate the proposed learning technique, we conduct extensive experiments on eighteen cross language sentiment classification tasks with four different languages. The empirical results demonstrate the efficacy of the proposed approach, and show it outperforms a number of related cross-lingual learning methods.

NeurIPS Conference 2013 Conference Paper

A Novel Two-Step Method for Cross Language Representation Learning

• Min Xiao
• Yuhong Guo

Cross language text classification is an important learning task in natural language processing. A critical challenge of cross language learning lies in that words of different languages are in disjoint feature spaces. In this paper, we propose a two-step representation learning method to bridge the feature spaces of different languages by exploiting a set of parallel bilingual documents. Specifically, we first formulate a matrix completion problem to produce a complete parallel document-term matrix for all documents in two languages, and then induce a cross-lingual document representation by applying latent semantic indexing on the obtained matrix. We use a projected gradient descent algorithm to solve the formulated matrix completion problem with convergence guarantees. The proposed approach is evaluated by conducting a set of experiments with cross language sentiment classification tasks on Amazon product reviews. The experimental results demonstrate that the proposed learning approach outperforms a number of comparison cross language representation learning methods, especially when the number of parallel bilingual documents is small.

IJCAI Conference 2013 Conference Paper

Active Learning with Multi-Label SVM Classification

• Xin Li
• Yuhong Guo

Multi-label classification, where each instance is assigned to multiple categories, is a prevalent problem in data analysis. However, annotations of multi-label instances are typically more timeconsuming or expensive to obtain than annotations of single-label instances. Though active learning has been widely studied on reducing labeling effort for single-label problems, current research on multi-label active learning remains in a preliminary state. In this paper, we first propose two novel multi-label active learning strategies, a max-margin prediction uncertainty strategy and a label cardinality inconsistency strategy, and then integrate them into an adaptive framework of multi-label active learning. Our empirical results on multiple multilabel data sets demonstrate the efficacy of the proposed active instance selection strategies and the integrated active learning approach.

AAAI Conference 2013 Conference Paper

Convex Subspace Representation Learning from Multi-View Data

• Yuhong Guo

Learning from multi-view data is important in many applications. In this paper, we propose a novel convex subspace representation learning method for unsupervised multi-view clustering. We first formulate the subspace learning with multiple views as a joint optimization problem with a common subspace representation matrix and a group sparsity inducing norm. By exploiting the properties of dual norms, we then show a convex min-max dual formulation with a sparsity inducing trace norm can be obtained. We develop a proximal bundle optimization algorithm to globally solve the minmax optimization problem. Our empirical study shows the proposed subspace representation learning method can effectively facilitate multi-view clustering and induce superior clustering results than alternative multiview clustering methods.

ICML Conference 2013 Conference Paper

Domain Adaptation for Sequence Labeling Tasks with a Probabilistic Language Adaptation Model

• Min Xiao 0004
• Yuhong Guo

In this paper, we propose to address the problem of domain adaptation for sequence labeling tasks via distributed representation learning by using a log-bilinear language adaptation model. The proposed neural probabilistic language model simultaneously models two different but related data distributions in the source and target domains based on induced distributed representations, which encode both generalizable and domain-specific latent features. We then use the learned dense real-valued representation as augmenting features for natural language processing systems. We empirically evaluate the proposed learning technique on WSJ and MEDLINE domains with POS tagging systems, and on WSJ and Brown corpora with syntactic chunking and name entity recognition systems. Our primary results show that the proposed domain adaptation method outperforms a number comparison methods for cross domain sequence labeling tasks.

IJCAI Conference 2013 Conference Paper

Probabilistic Multi-Label Classification with Sparse Feature Learning

• Yuhong Guo
• Wei Xue

Multi-label classification is a critical problem in many areas of data analysis such as image labeling and text categorization. In this paper we propose a probabilistic multi-label classification model based on novel sparse feature learning. By employing an individual sparsity inducing `1-norm and a group sparsity inducing `2, 1-norm, the proposed model has the capacity of capturing both label interdependencies and common predictive model structures. We formulate this sparse norm regularized learning problem as a non-smooth convex optimization problem, and develop a fast proximal gradient algorithm to solve it for an optimal solution. Our empirical study demonstrates the efficacy of the proposed method on a set of multi-label tasks given a limited number of labeled training instances.

NeurIPS Conference 2013 Conference Paper

Robust Transfer Principal Component Analysis with Rank Constraints

• Yuhong Guo

Principal component analysis (PCA), a well-established technique for data analysis and processing, provides a convenient form of dimensionality reduction that is effective for cleaning small Gaussian noises presented in the data. However, the applicability of standard principal component analysis in real scenarios is limited by its sensitivity to large errors. In this paper, we tackle the challenge problem of recovering data corrupted with errors of high magnitude by developing a novel robust transfer principal component analysis method. Our method is based on the assumption that useful information for the recovery of a corrupted data matrix can be gained from an uncorrupted related data matrix. Specifically, we formulate the data recovery problem as a joint robust principal component analysis problem on the two data matrices, with shared common principal components across matrices and individual principal components specific to each data matrix. The formulated optimization problem is a minimization problem over a convex objective function but with non-convex rank constraints. We develop an efficient proximal projected gradient descent algorithm to solve the proposed optimization problem with convergence guarantees. Our empirical results over image denoising tasks show the proposed method can effectively recover images with random large errors, and significantly outperform both standard PCA and robust PCA.

ICML Conference 2012 Conference Paper

Cross Language Text Classification via Subspace Co-regularized Multi-view Learning

• Yuhong Guo
• Min Xiao 0004

AAAI Conference 2012 Conference Paper

Learning SVM Classifiers with Indefinite Kernels

• Suicheng Gu
• Yuhong Guo

Recently, training support vector machines with indefinite kernels has attracted great attention in the machine learning community. In this paper, we tackle this problem by formulating a joint optimization model over SVM classifications and kernel principal component analysis. We first reformulate the kernel principal component analysis as a general kernel transformation framework, and then incorporate it into the SVM classification to formulate a joint optimization model. The proposed model has the advantage of making consistent kernel transformations over training and test samples. It can be used for both binary classification and multiclass classification problems. Our experimental results on both synthetic data sets and real world data sets show the proposed model can significantly outperform related approaches.

AAAI Conference 2012 Conference Paper

Semi-Supervised Kernel Matching for Domain Adaptation

• Min Xiao
• Yuhong Guo

In this paper, we propose a semi-supervised kernel matching method to address domain adaptation problems where the source distribution substantially differs from the target distribution. Specifically, we learn a prediction function on the labeled source data while mapping the target data points to similar source data points by matching the target kernel matrix to a submatrix of the source kernel matrix based on a Hilbert Schmidt Independence Criterion. We formulate this simultaneous learning and mapping process as a non-convex integer optimization problem and present a local minimization procedure for its relaxed continuous form. Our empirical results show the proposed kernel matching method significantly outperforms alternative methods on the task of across domain sentiment classification.

AAAI Conference 2011 Conference Paper

Adaptive Large Margin Training for Multilabel Classification

• Yuhong Guo
• Dale Schuurmans

Multilabel classification is a central problem in many areas of data analysis, including text and multimedia categorization, where individual data objects need to be assigned multiple labels. A key challenge in these tasks is to learn a classifier that can properly exploit label correlations without requiring exponential enumeration of label subsets during training or testing. We investigate novel loss functions for multilabel training within a large margin framework—identifying a simple alternative that yields improved generalization while still allowing efficient training. We furthermore show how covariances between the label models can be learned simultaneously with the classification model itself, in a jointly convex formulation, without compromising scalability. The resulting combination yields state of the art accuracy in multilabel webpage classification.

IJCAI Conference 2011 Conference Paper

Multi-Label Classification Using Conditional Dependency Networks

• Yuhong Guo
• Suicheng Gu

In this paper, we tackle the challenges of multi-label classification by developing a general conditional dependency network model. The proposed model is a cyclic directed graphical model, which provides an intuitive representation for the dependencies among multiple label variables, and a well integrated framework for efficient model training using binary classifiers and label predictions using Gibbs sampling inference. Our experiments show the proposed conditional model can effectively exploit the label dependency to improve multi-label classification performance.

NeurIPS Conference 2010 Conference Paper

Active Instance Sampling via Matrix Partition

• Yuhong Guo

Recently, batch-mode active learning has attracted a lot of attention. In this paper, we propose a novel batch-mode active learning approach that selects a batch of queries in each iteration by maximizing a natural form of mutual information criterion between the labeled and unlabeled instances. By employing a Gaussian process framework, this mutual information based instance selection problem can be formulated as a matrix partition problem. Although the matrix partition is an NP-hard combinatorial optimization problem, we show a good local solution can be obtained by exploiting an effective local optimization technique on the relaxed continuous optimization problem. The proposed active learning approach is independent of employed classification models. Our empirical studies show this approach can achieve comparable or superior performance to discriminative batch-mode active learning methods.

ECAI Conference 2010 Conference Paper

Prediction of Attributes and Links in Temporal Social Networks

• Vladimir Ouzienko
• Yuhong Guo
• Zoran Obradovic

The analysis of social networks often assumes the time invariant scenario while in practice node attributes and links in such networks often evolve over time. In this paper, we propose a new method to predict node attributes and links in temporal networks.

ECAI Conference 2010 Conference Paper

Regression Learning with Multiple Noisy Oracles

• Kosta Ristovski
• Debasish Das
• Vladimir Ouzienko
• Yuhong Guo
• Zoran Obradovic

In regression learning, it is often difficult to obtain the true values of the label variables, while multiple sources of noisy estimates of lower quality are readily available. To address this problem, we propose a new Bayesian approach that learns a regression model from data with noisy labels provided by multiple oracles. The proposed method provides closed form solution for model parameters and is applicable to both linear and nonlinear regression problems. In our experiments on synthetic and benchmark datasets this new regression model was consistently more accurate than a model trained with averaged estimates from multiple oracles as labels.

NeurIPS Conference 2008 Conference Paper

Supervised Exponential Family Principal Component Analysis via Convex Optimization

• Yuhong Guo

Recently, supervised dimensionality reduction has been gaining attention, owing to the realization that data labels are often available and strongly suggest important underlying structures in the data. In this paper, we present a novel convex supervised dimensionality reduction approach based on exponential family PCA and provide a simple but novel form to project new testing data into the embedded space. This convex approach successfully avoids the local optima of the EM learning. Moreover, by introducing a sample-based multinomial approximation to exponential family models, it avoids the limitation of the prevailing Gaussian assumptions of standard PCA, and produces a kernelized formulation for nonlinear supervised dimensionality reduction. A training algorithm is then devised based on a subgradient bundle method, whose scalability can be gained through a coordinate descent procedure. The advantage of our global optimization approach is demonstrated by empirical results over both synthetic and real data.

IJCAI Conference 2007 Conference Paper

• Yuhong Guo
• Russ Greiner

An active learning system will sequentially decide which unlabeled instance to label, with the goal of efficiently gathering the information necessary to produce a good classifier. Some such systems greedily select the next instance based only on properties of that instance and the few currently labeled points--- e. g. , selecting the one closest to the current classification boundary. Unfortunately, these approaches ignore the valuable information contained in the other unlabeled instances, which can help identify a good classifier much faster. For the previous approaches that do exploit this unlabeled data, this information is mostly used in a conservative way. One common property of the approaches in the literature is that the active learner sticks to one single query selection criterion in the whole process. We propose a system, MM+M, that selects the query instance that is able to provide the maximum conditional mutual information about the labels of the unlabeled instances, given the labeled data, in an optimistic way. This approach implicitly exploits the discriminative partition information contained in the unlabeled data. Instead of using one selection criterion, MM+M also employs a simple on-line method that changes its selection rule when it encounters an unexpected label. Our empirical results demonstrate that this new approach works effectively.

NeurIPS Conference 2007 Conference Paper

Convex Relaxations of Latent Variable Training

• Yuhong Guo
• Dale Schuurmans

We investigate a new, convex relaxation of an expectation-maximization (EM) variant that approximates a standard objective while eliminating local minima. First, a cautionary result is presented, showing that any convex relaxation of EM over hidden variables must give trivial results if any dependence on the missing values is retained. Although this appears to be a strong negative outcome, we then demonstrate how the problem can be bypassed by using equivalence relations instead of value assignments over hidden variables. In particular, we develop new algorithms for estimating exponential conditional models that only require equivalence relation information over the variable values. This reformulation leads to an exact expression for EM variants in a wide range of problems. We then develop a semidefinite relaxation that yields global training by eliminating local minima.

NeurIPS Conference 2007 Conference Paper

Discriminative Batch Mode Active Learning

• Yuhong Guo
• Dale Schuurmans

Active learning sequentially selects unlabeled instances to label with the goal of reducing the effort needed to learn a good classifier. Most previous studies in active learning have focused on selecting one unlabeled instance at one time while retraining in each iteration. However, single instance selection systems are unable to exploit a parallelized labeler when one is available. Recently a few batch mode active learning approaches have been proposed that select a set of most informative unlabeled instances in each iteration, guided by some heuristic scores. In this paper, we propose a discriminative batch mode active learning approach that formulates the instance selection task as a continuous optimization problem over auxiliary instance selection variables. The optimization is formuated to maximize the discriminative classification performance of the target classifier, while also taking the unlabeled data into account. Although the objective is not convex, we can manipulate a quasi-Newton method to obtain a good local solution. Our empirical studies on UCI datasets show that the proposed active learning is more effective than current state-of-the art batch mode active learning algorithms.

UAI Conference 2006 Conference Paper

Convex Structure Learning for Bayesian Networks: Polynomial Feature Selection and Approximate Ordering

• Yuhong Guo
• Dale Schuurmans

We present a new approach to learning the structure and parameters of a Bayesian network based on regularized estimation in an exponential family representation. Here we show that, given a fixed variable order, the optimal structure and parameters can be learned efficiently, even without restricting the size of the parent sets. We then consider the problem of optimizing the variable order for a given set of features. This is still a computationally hard problem, but we present a convex relaxation that yields an optimal 'soft' ordering in polynomial time. One novel aspect of the approach is that we do not perform a discrete search over DAG structures, nor over variable orders, but instead solve a continuous relaxation that can then be rounded to obtain a valid network structure. We conduct an experimental comparison against standard structure search procedures over standard objectives, which cope with local minima, and evaluate the advantages of using convex relaxations that reduce the effects of local minima.

AAAI Conference 2005 Conference Paper

Discriminative Model Selection for Belief Net Structures

• Yuhong Guo

Bayesian belief nets (BNs) are often used for classification tasks, typically to return the most likely class label for a specified instance. Many BN-learners, however, attempt to find the BN that maximizes a different objective function — viz. , likelihood, rather than classification accuracy — typically by first using some model selection criterion to identify an appropriate graphical structure, then finding good parameters for that structure. This paper considers a number of possible criteria for selecting the best structure, both generative (i. e. , based on likelihood; BIC, BDe) and discriminative (i. e. , Conditional BIC (CBIC), resubstitution Classification Error (CE) and Bias2 +Variance (BV) ). We empirically compare these criteria against a variety of different “correct BN structures”, both real-world and synthetic, over a range of complexities. We also explore different ways to set the parameters, dealing with two issues: (1) Should we seek the parameters that maximize likelihood versus the ones that maximize conditional likelihood? (2) Should we use (i) the entire training sample first to learn the best parameters and then to evaluate the models, versus (ii) only a partition for parameter estimation and another partition for evaluation (cross-validation)? Our results show that the discriminative BV model selection criterion is one of the best measures for identifying the optimal structure, while the discriminative CBIC performs poorly; that we should use the parameters that maximize likelihood; and that it is typically better to use cross-validation here.

IJCAI Conference 2005 Conference Paper

Learning Coordination Classifiers

• Yuhong Guo
• Russell Greiner
• Dale

We present a new approach to ensemble classification that requires learning only a single base classifier. The idea is to learn a classifier that simultaneously predicts pairs of test labels—as opposed to learning multiple predictors for single test labels— then coordinating the assignment of individual labels by propagating beliefs on a graph over the data. We argue that the approach is statistically well motivated, even for independent identically distributed (iid) data. In fact, we present experimental results that show improvements in classification accuracy over single-example classifiers, across a range of iid data sets and over a set of base classifiers. Like boosting, the technique increases representational capacity while controlling variance through a principled form of classifier combination.

UAI Conference 2005 Conference Paper

Maximum Margin Bayesian Networks

• Yuhong Guo
• Dana F. Wilkinson
• Dale Schuurmans

We consider the problem of learning Bayesian network classifiers that maximize the marginover a set of classification variables. We find that this problem is harder for Bayesian networks than for undirected graphical models like maximum margin Markov networks. The main difficulty is that the parameters in a Bayesian network must satisfy additional normalization constraints that an undirected graphical model need not respect. These additional constraints complicate the optimization task. Nevertheless, we derive an effective training algorithm that solves the maximum margin training problem for a range of Bayesian network topologies, and converges to an approximate solution for arbitrary network topologies. Experimental results show that the method can demonstrate improved generalization performance over Markov networks when the directed graphical structure encodes relevant knowledge. In practice, the training technique allows one to combine prior knowledge expressed as a directed (causal) model with state of the art discriminative learning methods.