Probabilistic sequence clustering with spectral learning

In this paper, we derive two novel learning algorithms for time series clustering; namely for learning mixtures of Markov Models and mixtures of Hidden Markov Models. Mixture models are special latent variable models that require the usage of local search heuristics such as Expectation Maximization (EM) algorithm, that can only provide locally optimal solutions. In contrast, we make use of the spectral learning algorithms, recently popularized in the machine learning community. Under mild assumptions, spectral learning algorithms are able to estimate the parameters in latent variable models by solving systems of equations via eigendecompositions of matrices or tensors of observable moments. As such, spectral methods can be viewed as an instance of the method of moments for parameter estimation, an alternative to maximum likelihood. The popularity stems from the fact that these methods provide a computationally cheap and local optima free alternative to EM. We conduct classification experiments on human action sequences extracted from videos, clustering experiments on motion capture data and network traffic data to illustrate the viability of our approach. We conclude that the spectral methods are a practical and useful alternative in terms of computational effort and solution quality to standard iterative techniques such as EM in several sequence clustering applications.     

Exploring context and content links in social media: a latent space method

Social media networks contain both content and context-specific information. Most existing methods work with either of the two for the purpose of multimedia mining and retrieval. In reality, both content and context information are rich sources of information for mining, and the full power of mining and processing algorithms can be realized only with the use of a combination of the two. This paper proposes a new algorithm which mines both context and content links in social media networks to discover the underlying latent semantic space. This mapping of the multimedia objects into latent feature vectors enables the use of any off-the-shelf multimedia retrieval algorithms. Compared to the state-of-the-art latent methods in multimedia analysis, this algorithm effectively solves the problem of sparse context links by mining the geometric structure underlying the content links between multimedia objects. Specifically for multimedia annotation, we show that an effective algorithm can be developed to directly construct annotation models by simultaneously leveraging both context and content information based on latent structure between correlated semantic concepts. We conduct experiments on the Flickr data set, which contains user tags linked with images. We illustrate the advantages of our approach over the state-of-the-art multimedia retrieval techniques.     

Flexible latent variable models for multi-task learning

Given multiple prediction problems such as regression or classification, we are interested in a joint inference framework that can effectively share information between tasks to improve the prediction accuracy, especially when the number of training examples per problem is small. In this paper we propose a probabilistic framework which can support a set of latent variable models for different multi-task learning scenarios. We show that the framework is a generalization of standard learning methods for single prediction problems and it can effectively model the shared structure among different prediction tasks. Furthermore, we present efficient algorithms for the empirical Bayes method as well as point estimation. Our experiments on both simulated datasets and real world classification datasets show the effectiveness of the proposed models in two evaluation settings: a standard multi-task learning setting and a transfer learning setting.     

Data augmentation strategies for the Bayesian spatial probit regression model

The well known latent variable representation of the Bayesian probit regression model due to Albert and Chib (1993) allows model fitting to be performed using a simple Gibbs sampler. In addition, various types of dependence among categorical outcomes not explained by covariate information can be accommodated in a straightforward manner as a result of this latent variable representation of the model. One example of this is the spatial probit regression model for spatially-referenced categorical outcomes. In this setting, commonly used covariance structures for describing residual spatial dependence in the normal linear model setting can be imbedded into the probit regression model. Capturing spatial dependence in this way, however, can negatively impact the performance of MCMC model-fitting algorithms, particularly in terms of mixing and sensitivity to starting values. To address these computational issues, we demonstrate how the non-identifiable spatial variance parameter can be used to create data augmentation MCMC algorithms. We compare the performance of several non-collapsed and partially collapsed data augmentation MCMC algorithms through a simulation study and an analysis of land cover data.     

Bridging systems theory and data science: A unifying review of dynamic latent variable analytics and process monitoring

This paper is concerned with data science and analytics as applied to data from dynamic systems for the purpose of monitoring, prediction, and inference. Collinearity is inevitable in industrial operation data. Therefore, we focus on latent variable methods that achieve dimension reduction and collinearity removal. We present a new dimension reduction expression of state space framework to unify dynamic latent variable analytics for process data, dynamic factor models for econometrics, subspace identification of multivariate dynamic systems, and machine learning algorithms for dynamic feature analysis. We unify or differentiate them in terms of model structure, objectives with constraints, and parsimony of parameterization. The Kalman filter theory in the latent space is used to give a system theory foundation to some empirical treatments in data analytics. We provide a unifying review of the connections among the dynamic latent variable methods, dynamic factor models, subspace identification methods, dynamic feature extractions, and their uses for prediction and process monitoring. Both unsupervised dynamic latent variable analytics and the supervised counterparts are reviewed. Illustrative examples are presented to show the similarities and differences among the analytics in extracting features for prediction and monitoring.     

融合物品相似性与流形正则化的矩阵分解推荐模型

矩阵分解是推荐系统中应用最为广泛的方法之一,但其对物品隐因子及其相似性学习不够充分.社会网络分析中认为相互连接的个体有一定共性,受此启发提出一个能够借助近邻关系有效学习物品隐因子及其相似性的矩阵分解推荐模型.首先基于评分矩阵对物品相似性计算进行改良,综合同一用户和相似用户的评分共现信息对物品信息建模;然后通过构建相似性优化和流形局部保持正则化项,使物品相似性作用在矩阵分解中,从而充分学习物品隐因子特征及其相似性;最后根据用户和物品隐因子矩阵计算推荐指数.在公开数据集上的实验结果表明,通过流形正则化技术将改良的物品相似性作用在矩阵分解中,可以有效提升推荐效果.     

Shared Kernel Information Embedding for discriminative inference

Latent variable models, such as the GPLVM and related methods, help mitigate overfitting when learning from small or moderately sized training sets. Nevertheless, existing methods suffer from several problems: 1) complexity, 2) the lack of explicit mappings to and from the latent space, 3) an inability to cope with multimodality, and 4) the lack of a well-defined density over the latent space. We propose an LVM called the Kernel Information Embedding (KIE) that defines a coherent joint density over the input and a learned latent space. Learning is quadratic, and it works well on small data sets. We also introduce a generalization, the shared KIE (sKIE), that allows us to model multiple input spaces (e.g., image features and poses) using a single, shared latent representation. KIE and sKIE permit missing data during inference and partially labeled data during learning. We show that with data sets too large to learn a coherent global model, one can use the sKIE to learn local online models. We use sKIE for human pose inference.     

ConvNet frameworks for multi-modal fake news detection

An upsurge of false information revolves around the internet. Social media and websites are flooded with unverified news posts. These posts are comprised of text, images, audio, and videos. There is a requirement for a system that detects fake content in multiple data modalities. We have seen a considerable amount of research on classification techniques for textual fake news detection, while frameworks dedicated to visual fake news detection are very few. We explored the state-of-the-art methods using deep networks such as CNNs and RNNs for multi-modal online information credibility analysis. They show rapid improvement in classification tasks without requiring pre-processing. To aid the ongoing research over fake news detection using CNN models, we build textual and visual modules to analyze their performances over multi-modal datasets. We exploit latent features present inside text and images using layers of convolutions. We see how well these convolutional neural networks perform classification when provided with only latent features and analyze what type of images are needed to be fed to perform efficient fake news detection. We propose a multi-modal Coupled ConvNet architecture that fuses both the data modules and efficiently classifies online news depending on its textual and visual content. We thence offer a comparative analysis of the results of all the models utilized over three datasets. The proposed architecture outperforms various state-of-the-art methods for fake news detection with considerably high accuracies.

     

Topic modelling of clickthrough data in image search

In this paper we explore the benefits of latent variable modelling of clickthrough data in the domain of image retrieval. Clicks in image search logs are regarded as implicit relevance judgements that express both user intent and important relations between selected documents. We posit that clickthrough data contains hidden topics and can be used to infer a lower dimensional latent space that can be subsequently employed to improve various aspects of the retrieval system. We use a subset of a clickthrough corpus from the image search portal of a news agency to evaluate several popular latent variable models in terms of their ability to model topics underlying queries. We demonstrate that latent variable modelling reveals underlying structure in clickthrough data and our results show that computing document similarities in the latent space improves retrieval effectiveness compared to computing similarities in the original query space. These results are compared with baselines using visual and textual features. We show performance substantially better than the visual baseline, which indicates that content-based image retrieval systems that do not exploit query logs could improve recall and precision by taking this historical data into account.     

Combining Statistical Language Models via the Latent Maximum Entropy Principle

We present a unified probabilistic framework for statistical language modeling which can simultaneously incorporate various aspects of natural language, such as local word interaction, syntactic structure and semantic document information. Our approach is based on a recent statistical inference principle we have proposed—the latent maximum entropy principle—which allows relationships over hidden features to be effectively captured in a unified model. Our work extends previous research on maximum entropy methods for language modeling, which only allow observed features to be modeled. The ability to conveniently incorporate hidden variables allows us to extend the expressiveness of language models while alleviating the necessity of pre-processing the data to obtain explicitly observed features. We describe efficient algorithms for marginalization, inference and normalization in our extended models. We then use these techniques to combine two standard forms of language models: local lexical models (Markov N-gram models) and global document-level semantic models (probabilistic latent semantic analysis). Our experimental results on the Wall Street Journal corpus show that we obtain a 18.5% reduction in perplexity compared to the baseline tri-gram model with Good-Turing smoothing.Editors: Dan Roth and Pascale Fung     

Unsupervised Object Discovery: A Comparison

The goal of this paper is to evaluate and compare models and methods for learning to recognize basic entities in images in an unsupervised setting. In other words, we want to discover the objects present in the images by analyzing unlabeled data and searching for re-occurring patterns. We experiment with various baseline methods, methods based on latent variable models, as well as spectral clustering methods. The results are presented and compared both on subsets of Caltech256 and MSRC2, data sets that are larger and more challenging and that include more object classes than what has previously been reported in the literature. A rigorous framework for evaluating unsupervised object discovery methods is proposed.     

Emphasizing Essential Words for Sentiment Classification Based on Recurrent Neural Networks

With the explosion of online communication and publication, texts become obtainable via forums, chat messages, blogs, book reviews and movie reviews. Usually, these texts are much short and noisy without sufficient statistical signals and enough information for a good semantic analysis. Traditional natural language processing methods such as Bow-of-Word (BOW) based probabilistic latent semantic models fail to achieve high performance due to the short text environment. Recent researches have focused on the correlations between words, i.e., term dependencies, which could be helpful for mining latent semantics hidden in short texts and help people to understand them. Long short-term memory (LSTM) network can capture term dependencies and is able to remember the information for long periods of time. LSTM has been widely used and has obtained promising results in variants of problems of understanding latent semantics of texts. At the same time, by analyzing the texts, we find that a number of keywords contribute greatly to the semantics of the texts. In this paper, we establish a keyword vocabulary and propose an LSTM-based model that is sensitive to the words in the vocabulary; hence, the keywords leverage the semantics of the full document. The proposed model is evaluated in a short-text sentiment analysis task on two datasets: IMDB and SemEval-2016, respectively. Experimental results demonstrate that our model outperforms the baseline LSTM by 1%~2% in terms of accuracy and is effective with significant performance enhancement over several non-recurrent neural network latent semantic models (especially in dealing with short texts). We also incorporate the idea into a variant of LSTM named the gated recurrent unit (GRU) model and achieve good performance, which proves that our method is general enough to improve different deep learning models.     

Leveraging clustering to improve collaborative filtering

Extensive work on matrix factorization (MF) techniques have been done recently as they provide accurate rating prediction models in recommendation systems. Additional extensions, such as neighbour-aware models, have been shown to improve rating prediction further. However, these models often suffer from a long computation time. In this paper, we propose a novel method that applies clustering algorithms to the latent vectors of users and items. Our method can capture the common interests between the cluster of users and the cluster of items in a latent space. A matrix factorization technique is then applied to this cluster-level rating matrix to predict the future cluster-level interests. We then aggregate the traditional user-item rating predictions with our cluster-level rating predictions to improve the rating prediction accuracy. Our method is a general “wrapper” that can be applied to all collaborative filtering methods. In our experiments, we show that our new approach, when applied to a variety of existing matrix factorization techniques, improves their rating predictions and also results in better rating predictions for cold-start users. Above all, in this paper we show that better quality and more quantity of these clusters achieve a better rating prediction accuracy.     

Optimal flight scheduling models for cargo airlines under alliances

In this paper we develop several coordinated scheduling models combining airport selection, fleet routing and timetable setting, in order to help airlines solve for the most satisfactory cargo fleet routes and timetables when they enter into alliances. We employ network flow techniques to construct the models, which are formulated as a multiple commodity network flow problem and can be solved using a mathematical programming solver. To evaluate the models, we perform numerical tests based on real operating data from two Taiwan airlines. The preliminary results are good, showing that the models would be useful for airline alliances.     

Efficient direct sampling MCEM algorithm for latent variable models with binary responses

While latent variable models have been successfully applied in many fields and underpin various modeling techniques, their ability to incorporate categorical responses is hindered due to the lack of accurate and efficient estimation methods. Approximation procedures, such as penalized quasi-likelihood, are computationally efficient, but the resulting estimators can be seriously biased for binary responses. Gauss-Hermite quadrature and Markov Chain Monte Carlo (MCMC) integration based methods can yield more accurate estimation, but they are computationally much more intensive. Estimation methods that can achieve both computational efficiency and estimation accuracy are still under development. This paper proposes an efficient direct sampling based Monte Carlo EM algorithm (DSMCEM) for latent variable models with binary responses. Mixed effects and item factor analysis models with binary responses are used to illustrate this algorithm. Results from two simulation studies and a real data example suggest that, as compared with MCMC based EM, DSMCEM can significantly improve computational efficiency as well as produce equally accurate parameter estimates. Other aspects and extensions of the algorithm are discussed.     

Filter-Based Mean-Field Inference for Random Fields with Higher-Order Terms and Product Label-Spaces

Recently, a number of cross bilateral filtering methods have been proposed for solving multi-label problems in computer vision, such as stereo, optical flow and object class segmentation that show an order of magnitude improvement in speed over previous methods. These methods have achieved good results despite using models with only unary and/or pairwise terms. However, previous work has shown the value of using models with higher-order terms e.g. to represent label consistency over large regions, or global co-occurrence relations. We show how these higher-order terms can be formulated such that filter-based inference remains possible. We demonstrate our techniques on joint stereo and object labelling problems, as well as object class segmentation, showing in addition for joint object-stereo labelling how our method provides an efficient approach to inference in product label-spaces. We show that we are able to speed up inference in these models around 10–30 times with respect to competing graph-cut/move-making methods, as well as maintaining or improving accuracy in all cases. We show results on PascalVOC-10 for object class segmentation, and Leuven for joint object-stereo labelling.     

A simple index for assessing fuel moisture content

Assessing fuel moisture content to within a reasonable degree of accuracy is an important part of wildland fire management. In this paper we introduce a fuel moisture index that provides a simple and intuitive method for assessing fuel moisture content. The method can be quickly and easily applied in a field setting to provide a dimensionless measure of fuel moisture content. We compare the index with predictions from several models for fuel moisture content and conclude that it provides an equivalent measure of fuel moisture content for a number of fuel types. We go on to briefly discuss how the index could be used to construct a simple and intuitive fire danger index.     

LDA-CF: 一种混合协同过滤方法

推荐系统是一种克服信息过载的重要工具,其中最流行的方法是协同过滤。该文提出一种结合潜在因素模型和邻域方法的混合协同过滤方法LDA-CF。我们首先将评分矩阵转换成伪文档集合,使用LDA(Latent Dirichlet Allocation)主题模型发现用户和物品潜在因素向量;然后在低维潜在因素空间计算用户和物品相似度;最后采用邻域方法预测未知评分。在MovieLens 100k数据集上的实验表明: 在评分预测任务中,LDA-CF取得的MAE性能指标优于传统的邻域方法。因此,LDA可以有效地从评分矩阵中发现对计算相似度十分有用的用户和物品低维特征表示,在一定程度上缓解了数据稀疏问题。     

A differential privacy framework for matrix factorization recommender systems

Recommender systems rely on personal information about user behavior for the recommendation generation purposes. Thus, they inherently have the potential to hamper user privacy and disclose sensitive information. Several works studied how neighborhood-based recommendation methods can incorporate user privacy protection. However, privacy preserving latent factor models, in particular, those represented by matrix factorization techniques, the state-of-the-art in recommender systems, have received little attention. In this paper, we address the problem of privacy preserving matrix factorization by utilizing differential privacy, a rigorous and provable approach to privacy in statistical databases. We propose a generic framework and evaluate several ways, in which differential privacy can be applied to matrix factorization. By doing so, we specifically address the privacy-accuracy trade-off offered by each of the algorithms. We show that, of all the algorithms considered, input perturbation results in the best recommendation accuracy, while guaranteeing a solid level of privacy protection against attacks that aim to gain knowledge about either specific user ratings or even the existence of these ratings. Our analysis additionally highlights the system aspects that should be addressed when applying differential privacy in practice, and when considering potential privacy preserving solutions.     

Ancillarity-sufficiency interweaving strategy (ASIS) for boosting MCMC estimation of stochastic volatility models

Bayesian inference for stochastic volatility models using MCMC methods highly depends on actual parameter values in terms of sampling efficiency. While draws from the posterior utilizing the standard centered parameterization break down when the volatility of volatility parameter in the latent state equation is small, non-centered versions of the model show deficiencies for highly persistent latent variable series. The novel approach of ancillarity-sufficiency interweaving has recently been shown to aid in overcoming these issues for a broad class of multilevel models. It is demonstrated how such an interweaving strategy can be applied to stochastic volatility models in order to greatly improve sampling efficiency for all parameters and throughout the entire parameter range. Moreover, this method of “combining best of different worlds” allows for inference for parameter constellations that have previously been infeasible to estimate without the need to select a particular parameterization beforehand.