The Gartner's 2014 Hype Cycle released last August moves Big Data technology from the Peak of Inflated Expectations to the beginning of the Trough of Disillusionment when interest starts to wane as reality does not live up to previous promises. As the hype is starting to dissipate it is worth asking what Big Data (however defined) means from a scientific perspective: Did the emergence of gigantic corpora exposed the limits of classical information retrieval and data mining and led to new concepts and challenges, the way say, the study of electromagnetism showed the limits of Newtonian mechanics and led to Relativity Theory, or is it all just "sound and fury, signifying nothing", simply a matter of scaling up well understood technologies? To answer this question, we have assembled a distinguished panel of eminent scientists, from both Industry and Academia: Lada Adamic (Facebook), Michael Franklin (University of California at Berkeley), Maarten de Rijke (University of Amsterdam), Eric Xing (Carnegie Mellon University), and Kai Yu (Baidu) will share their point of view and take questions from the moderator and the audience.

Recently two lines of image retrieval algorithms demonstrate excellent scalability: 1) local features indexed by a vocabulary tree, and 2) holistic features indexed by compact hashing codes. Although both of them are able to search visually similar images effectively, their retrieval precision may vary dramatically among queries. Therefore, combining these two types of methods is expected to further enhance the retrieval precision. However, the feature characteristics and the algorithmic procedures of these methods are dramatically different, which is very challenging for the feature-level fusion. This motivates us to investigate how to fuse the ordered retrieval sets, i.e., the ranks of images, given by multiple retrieval methods, to boost the retrieval precision without sacrificing their scalability. In this paper, we model retrieval ranks as graphs of candidate images and propose a graph-based query specific fusion approach, where multiple graphs are merged and reranked by conducting a link analysis on a fused graph. The retrieval quality of an individual method is measured on-the-fly by assessing the consistency of the top candidates' nearest neighborhoods. Hence, it is capable of adaptively integrating the strengths of the retrieval methods using local or holistic features for different query images. This proposed method does not need any supervision, has few parameters, and is easy to implement. Extensive and thorough experiments have been conducted on four public datasets, i.e., the UKbench, Corel-5K, Holidays and the large-scale San Francisco Landmarks datasets. Our proposed method has achieved very competitive performance, including state-of-the-art results on several data sets, e.g., the N-S score 3.83 for UKbench.

In the past 30 years, tremendous progress has been achieved in building effective shallow classification models. Despite the success, we come to realize that, for many applications, the key bottleneck is not the qualify of classifiers but that of features. Not being able to automatically get useful features has become the main limitation for shallow models. Since 2006, learning high-level features using deep architectures from raw data has become a huge wave of new learning paradigms. In recent two years, deep learning has made many performance breakthroughs, for example, in the areas of image understanding and speech recognition. In this talk, I will walk through some of the latest technology advances of deep learning within Baidu, and discuss the main challenges, e.g., developing effective models for various applications, and scaling up the model training using many GPUs. In the end of the talk I will discuss what might be interesting future directions.

We propose and analyze a novel framework for learning sparse representations, based on two statistical techniques: kernel smoothing and marginal regression. The proposed approach provides a flexible framework for incorporating feature similarity or temporal information present in data sets, via nonparametric kernel smoothing. We provide generalization bounds for dictionary learning using smooth sparse coding and show how the sample complexity depends on the L1 norm of kernel function used. Furthermore, we propose using marginal regression for obtaining sparse codes, which significantly improves the speed and allows one to scale to large dictionary sizes easily. We demonstrate the advantages of the proposed approach, both in terms of accuracy and speed by extensive experimentation on several real data sets. In addition, we demonstrate how the proposed approach could be used for improving semisupervised sparse coding.

We consider the automated recognition of human actions in surveillance videos. Most current methods build classifiers based on complex handcrafted features computed from the raw inputs. Convolutional neural networks (CNNs) are a type of deep model that can act directly on the raw inputs. However, such models are currently limited to handling 2D inputs. In this paper, we develop a novel 3D CNN model for action recognition. This model extracts features from both the spatial and the temporal dimensions by performing 3D convolutions, thereby capturing the motion information encoded in multiple adjacent frames. The developed model generates multiple channels of information from the input frames, and the final feature representation combines information from all channels. To further boost the performance, we propose regularizing the outputs with high-level features and combining the predictions of a variety of different models. We apply the developed models to recognize human actions in the real-world environment of airport surveillance videos, and they achieve superior performance in comparison to baseline methods.

We apply salient feature detection and tracking in videos to simulate fixations and smooth pursuit in human vision. With tracked sequences as input, a hierarchical network of modules learns invariant features using a temporal slowness constraint. The network encodes invariance which are increasingly complex with hierarchy. Although learned from videos, our features are spatial instead of spatial-temporal, and well suited for extracting features from still images. We applied our features to four datasets (COIL-100, Caltech 101, STL-10, PubFig), and observe a consistent improvement of 4% to 5% in classification accuracy. With this approach, we achieve state-of-the-art recognition accuracy 61% on STL-10 dataset.

Most research efforts on image classification so far have been focused on medium-scale datasets, which are often defined as datasets that can fit into the memory of a desktop (typically 4G~48G). There are two main reasons for the limited effort on large-scale image classification. First, until the emergence of ImageNet dataset, there was almost no publicly available large-scale benchmark data for image classification. This is mostly because class labels are expensive to obtain. Second, large-scale classification is hard because it poses more challenges than its medium-scale counterparts. A key challenge is how to achieve efficiency in both feature extraction and classifier training without compromising performance. This paper is to show how we address this challenge using ImageNet dataset as an example. For feature extraction, we develop a Hadoop scheme that performs feature extraction in parallel using hundreds of mappers. This allows us to extract fairly sophisticated features (with dimensions being hundreds of thousands) on 1.2 million images within one day. For SVM training, we develop a parallel averaging stochastic gradient descent (ASGD) algorithm for training one-against-all 1000-class SVM classifiers. The ASGD algorithm is capable of dealing with terabytes of training data and converges very fast-typically 5 epochs are sufficient. As a result, we achieve state-of-the-art performance on the ImageNet 1000-class classification, i.e., 52.9% in classification accuracy and 71.8% in top 5 hit rate.

Visual recognition systems for videos using statistical learning models often show degraded performance when being deployed to a real-world environment, primarily due to the fact that training data can hardly cover sufficient variations in reality. To alleviate this issue, we propose to utilize the object correspondences in successive frames as weak supervision to adapt visual recognition models, which is particularly suitable for human profile recognition. Specifically, we substantialize this new strategy on an advanced convolutional neural network (CNN) based system to estimate human gender, age, and race. We enforce the system to output consistent and stable results on face images from the same trajectories in videos by using incremental stochastic training. Our baseline system already achieves competitive performance on gender and age estimation as compared to the state-of-the-art algorithms on the FG-NET database. Further, on two new video datasets containing about 900 persons, the proposed supervision of correspondences improves the estimation accuracy by a large margin over the baseline.

We present a method for learning image representations using a two-layer sparse coding scheme at the pixel level. The first layer encodes local patches of an image. After pooling within local regions, the first layer codes are then passed to the second layer, which jointly encodes signals from the region. Unlike traditional sparse coding methods that encode local patches independently, this approach accounts for high-order dependency among patterns in a local image neighborhood. We develop algorithms for data encoding and codebook learning, and show in experiments that the method leads to more invariant and discriminative image representations. The algorithm gives excellent results for hand-written digit recognition on MNIST and object recognition on the Caltech101 benchmark. This marks the first time that such accuracies have been achieved using automatically learned features from the pixel level, rather than using hand-designed descriptors.

In this paper we address the problem of image retrieval from millions of database images. We improve the vocabulary tree based approach by introducing contextual weighting of local features in both descriptor and spatial domains. Specifically, we propose to incorporate efficient statistics of neighbor descriptors both on the vocabulary tree and in the image spatial domain into the retrieval. These contextual cues substantially enhance the discriminative power of individual local features with very small computational overhead. We have conducted extensive experiments on benchmark datasets, i.e., the UKbench, Holidays, and our new Mobile dataset, which show that our method reaches state-of-the-art performance with much less computation. Furthermore, the proposed method demonstrates excellent scalability in terms of both retrieval accuracy and efficiency on large-scale experiments using 1.26 million images from the ImageNet database as distractors.

Recognition of clothing categories from videos is appealing to emerging applications such as intelligent customer profile analysis and computer-aided fashion design. This paper presents a complete system to tag clothing categories in real-time, which addresses some practical complications in surveillance videos. Specifically, we take advantage of face detection and tracking to locate human figures and develop an efficient clothing segmentation method utilizing Voronoi images to select seeds for region growing. We compare clothing representations combining color histograms and 3 different texture descriptors. Evaluated on a video dataset with 937 persons and 25441 cloth instances, the system demonstrates promising results in recognizing 8 clothing categories.

The traditional SPM approach based on bag-of-features (BoF) requires nonlinear classifiers to achieve good image classification performance. This paper presents a simple but effective coding scheme called Locality-constrained Linear Coding (LLC) in place of the VQ coding in traditional SPM. LLC utilizes the locality constraints to project each descriptor into its local-coordinate system, and the projected coordinates are integrated by max pooling to generate the final representation. With linear classifier, the proposed approach performs remarkably better than the traditional nonlinear SPM, achieving state-of-the-art performance on several benchmarks. Compared with the sparse coding strategy [22], the objective function used by LLC has an analytical solution. In addition, the paper proposes a fast approximated LLC method by first performing a K-nearest-neighbor search and then solving a constrained least square fitting problem, bearing computational complexity of O(M + K2). Hence even with very large codebooks, our system can still process multiple frames per second. This efficiency significantly adds to the practical values of LLC for real applications.

In this paper, we propose a novel supervised hierarchical sparse coding model based on local image descriptors for classification tasks. The supervised dictionary training is performed via back-projection, by minimizing the training error of classifying the image level features, which are extracted by max pooling over the sparse codes within a spatial pyramid. Such a max pooling procedure across multiple spatial scales offer the model translation invariant properties, similar to the Convolutional Neural Network (CNN). Experiments show that our supervised dictionary improves the performance of the proposed model significantly over the unsupervised dictionary, leading to state-of-the-art performance on diverse image databases. Further more, our supervised model targets learning linear features, implying its great potential in handling large scale datasets in real applications.

This paper proposes a principled extension of the traditional single-layer flat sparse coding scheme, where a two-layer coding scheme is derived based on theoretical analysis of nonlinear functional approximation that extends recent results for local coordinate coding. The two-layer approach can be easily generalized to deeper structures in a hierarchical multiple-layer manner. Empirically, it is shown that the deep coding approach yields improved performance in benchmark datasets.

Gene expression data usually contain a large number of genes but a small number of samples. Feature selection for gene expression data aims at finding a set of genes that best discriminate biological samples of different types. Using machine learning techniques, traditional gene selection based on empirical mutual information suffers the data sparseness issue due to the small number of samples. To overcome the sparseness issue, we propose a model-based approach to estimate the entropy of class variables on the model, instead of on the data themselves. Here, we use multivariate normal distributions to fit the data, because multivariate normal distributions have maximum entropy among all real-valued distributions with a specified mean and standard deviation and are widely used to approximate various distributions. Given that the data follow a multivariate normal distribution, since the conditional distribution of class variables given the selected features is a normal distribution, its entropy can be computed with the log-determinant of its covariance matrix. Because of the large number of genes, the computation of all possible log-determinants is not efficient. We propose several algorithms to largely reduce the computational cost. The experiments on seven gene data sets and the comparison with other five approaches show the accuracy of the multivariate Gaussian generative model for feature selection, and the efficiency of our algorithms.

Recently SVMs using spatial pyramid matching (SPM) kernel have been highly successful in image classification. Despite its popularity, these nonlinear SVMs have a complexity O(n2∼ n3) in training and O(n) in testing, where n is the training size, implying that it is nontrivial to scaleup the algorithms to handlemore than thousands of training images. In this paper we develop an extension of the SPM method, by generalizing vector quantization to sparse coding followed by multi-scale spatial max pooling, and propose a linear SPM kernel based on SIFT sparse codes. This new approach remarkably reduces the complexity of SVMs to O(n) in training and a constant in testing. In a number of image categorization experiments, we find that, in terms of classification accuracy, the suggested linear SPM based on sparse coding of SIFT descriptors always significantly outperforms the linear SPM kernel on histograms, and is even better than the nonlinear SPM kernels, leading to state-of-the-art performance on several benchmarks by using a single type of descriptors.

No abstract available.

A nonparametric model is introduced that allows multiple related regression tasks to take inputs from a common data space. Traditional transfer learning models can be inappropriate if the dependence among the outputs cannot be fully resolved by known input-specific and task-specific predictors. The proposed model treats such output responses as conditionally independent, given known predictors and appropriate unobserved random effects. The model is nonparametric in the sense that the dimensionality of random effects is not specified a priori but is instead determined from data. An approach to estimating the model is presented uses an EM algorithm that is efficient on a very large scale collaborative prediction problem. The obtained prediction accuracy is competitive with state-of-the-art results.

Event detection plays an essential role in video content analysis and remains a challenging open problem. In particular, the study on detecting human-related video events in complex scenes with both a crowd of people and dynamic motion is still limited. In this paper, we investigate detecting video events that involve elementary human actions, e.g. making cellphone call, putting an object down, and pointing to something, in complex scenes using a novel spatio-temporal descriptor based approach. A new spatio-temporal descriptor, which temporally integrates the statistics of a set of response maps of low-level features, e.g. image gradients and optical flows, in a space-time cube, is proposed to capture the characteristics of actions in terms of their appearance and motion patterns. Based on this kind of descriptors, the bag-of-words method is utilized to describe a human figure as a concise feature vector. Then, these features are employed to train SVM classifiers at multiple spatial pyramid levels to distinguish different actions. Finally, a Gaussian kernel based temporal filtering is conducted to segment the sequences of events from a video stream taking account of the temporal consistency of actions. The proposed approach is capable of tolerating spatial layout variations and local deformations of human actions due to diverse view angles and rough human figure alignment in complex scenes. Extensive experiments on the 50-hour video dataset of TRECVid 2008 event detection task demonstrate that our approach outperforms the well-known SIFT descriptor based methods and effectively detects video events in challenging real-world conditions.

This paper introduces a new method for semi-supervised learning on high dimensional nonlinear manifolds, which includes a phase of unsupervised basis learning and a phase of supervised function learning. The learned bases provide a set of anchor points to form a local coordinate system, such that each data point x on the manifold can be locally approximated by a linear combination of its nearby anchor points, and the linear weights become its local coordinate coding. We show that a high dimensional nonlinear function can be approximated by a global linear function with respect to this coding scheme, and the approximation quality is ensured by the locality of such coding. The method turns a difficult nonlinear learning problem into a simple global linear learning problem, which overcomes some drawbacks of traditional local learning methods.

With the sheer growth of online user data, it becomes challenging to develop preference learning algorithms that are sufficiently flexible in modeling but also affordable in computation. In this paper we develop nonparametric matrix factorization methods by allowing the latent factors of two low-rank matrix factorization methods, the singular value decomposition (SVD) and probabilistic principal component analysis (pPCA), to be data-driven, with the dimensionality increasing with data size. We show that the formulations of the two nonparametric models are very similar, and their optimizations share similar procedures. Compared to traditional parametric low-rank methods, nonparametric models are appealing for their flexibility in modeling complex data dependencies. However, this modeling advantage comes at a computational price--it is highly challenging to scale them to large-scale problems, hampering their application to applications such as collaborative filtering. In this paper we introduce novel optimization algorithms, which are simple to implement, which allow learning both nonparametric matrix factorization models to be highly efficient on large-scale problems. Our experiments on EachMovie and Netflix, the two largest public benchmarks to date, demonstrate that the nonparametric models make more accurate predictions of user ratings, and are computationally comparable or sometimes even faster in training, in comparison with previous state-of-the-art parametric matrix factorization models.

In this paper we focus on training deep neural networks for visual recognition tasks. One challenge is the lack of an informative regularization on the network parameters, to imply a meaningful control on the computed function. We propose a training strategy that takes advantage of kernel methods, where an existing kernel function represents useful prior knowledge about the learning task of interest. We derive an efficient algorithm using stochastic gradient descent, and demonstrate very positive results in a wide range of visual recognition tasks.

Stochastic relational models provide a rich family of choices for learning and predicting dyadic data between two sets of entities. It generalizes matrix factorization to a supervised learning problem that utilizes attributes of objects in a hierarchical Bayesian framework. Previously empirical Bayesian inference was applied, which is however not scalable when the size of either object sets becomes tens of thousands. In this paper, we introduce a Markov chain Monte Carlo (MCMC) algorithm to scale the model to very large-scale dyadic data. Both superior scalability and predictive accuracy are demonstrated on a collaborative filtering problem, which involves tens of thousands users and a half million items.

In this paper we propose a non-greedy active learning method for text categorization using least-squares support vector machines (LSSVM). Our work is based on transductive experimental design (TED), an active learning formulation that effectively explores the information of unlabeled data. Despite its appealing properties, the optimization problem is however NP-hard and thus--like most of other active learning methods--a greedy sequential strategy to select one data example after another was suggested to find a suboptimum. In this paper we formulate the problem into a continuous optimization problem and prove its convexity, meaning that a set of data examples can be selected with a guarantee of global optimum. We also develop an iterative algorithm to efficiently solve the optimization problem, which turns out to be very easy-to-implement. Our text categorization experiments on two text corpora empirically demonstrated that the new active learning algorithm outperforms the sequential greedy algorithm, and is promising for active text categorization applications.

The Web offers rich relational data with different semantics. In this paper, we address the problem of document recommendation in a digital library, where the documents in question are networked by citations and are associated with other entities by various relations. Due to the sparsity of a single graph and noise in graph construction, we propose a new method for combining multiple graphs to measure document similarities, where different factorization strategies are used based on the nature of different graphs. In particular, the new method seeks a single low-dimensional embedding of documents that captures their relative similarities in a latent space. Based on the obtained embedding, a new recommendation framework is developed using semi-supervised learning on graphs. In addition, we address the scalability issue and propose an incremental algorithm. The new incremental method significantly improves the efficiency by calculating the embedding for new incoming documents only. The new batch and incremental methods are evaluated on two real world datasets prepared from CiteSeer. Experiments demonstrate significant quality improvement for our batch method and significant efficiency improvement with tolerable quality loss for our incremental method.

This paper presents a Local Learning Projection (LLP) approach for linear dimensionality reduction. We first point out that the well known Principal Component Analysis (PCA) essentially seeks the projection that has the minimal global estimation error. Then we propose a dimensionality reduction algorithm that leads to the projection with the minimal local estimation error, and elucidate its advantages for classification tasks. We also indicate that LLP keeps the local information in the sense that the projection value of each point can be well estimated based on its neighbors and their projection values. Experimental results are provided to validate the effectiveness of the proposed algorithm.

Most current multi-task learning frameworks ignore the robustness issue, which means that the presence of "outlier" tasks may greatly reduce overall system performance. We introduce a robust framework for Bayesian multitask learning, t-processes (TP), which are a generalization of Gaussian processes (GP) for multi-task learning. TP allows the system to effectively distinguish good tasks from noisy or outlier tasks. Experiments show that TP not only improves overall system performance, but can also serve as an indicator for the "informativeness" of different tasks.

In this paper we develop a Gaussian process (GP) framework to model a collection of reciprocal random variables defined on the \emph{edges} of a network. We show how to construct GP priors, i.e.,~covariance functions, on the edges of directed, undirected, and bipartite graphs. The model suggests an intimate connection between \emph{link prediction} and \emph{transfer learning}, which were traditionally considered two separate research topics. Though a straightforward GP inference has a very high complexity, we develop an efficient learning algorithm that can handle a large number of observations. The experimental results on several real-world data sets verify superior learning capacity.

It is becoming increasingly important to learn from a partially-observed random matrix and predict its missing elements. We assume that the entire matrix is a single sample drawn from a matrix-variate t distribution and suggest a matrix-variate t model (MVTM) to predict those missing elements. We show that MVTM generalizes a range of known probabilistic models, and automatically performs model selection to encourage sparse predictive models. Due to the non-conjugacy of its prior, it is difficult to make predictions by computing the mode or mean of the posterior distribution. We suggest an optimization method that sequentially minimizes a convex upper-bound of the log-likelihood, which is very efficient and scalable. The experiments on a toy data and EachMovie dataset show a good predictive accuracy of the model.

The world wide web contains rich textual contents that areinterconnected via complex hyperlinks. This huge database violates the assumption held by most of conventional statistical methods that each web page is considered as an independent and identical sample. It is thus difficult to apply traditional mining or learning methods for solving web mining problems, e.g., web page classification, by exploiting both the content and the link structure. The research in this direction has recently received considerable attention but are still in an early stage. Though a few methods exploit both the link structure or the content information, some of them combine the only authority information with the content information, and the others first decompose the link structure into hub and authority features, then apply them as additional document features. Being practically attractive for its great simplicity, this paper aims to design an algorithm that exploits both the content and linkage information, by carrying out a joint factorization on both the linkage adjacency matrix and the document-term matrix, and derives a new representation for web pages in a low-dimensional factor space, without explicitly separating them as content, hub or authority factors. Further analysis can be performed based on the compact representation of web pages. In the experiments, the proposed method is compared with state-of-the-art methods and demonstrates an excellent accuracy in hypertext classification on the WebKB and Cora benchmarks.

This paper considers the problem of selecting the most informative experiments x to get measurements y for learning a regression model y = f(x). We propose a novel and simple concept for active learning, transductive experimental design, that explores available unmeasured experiments (i.e., unlabeled data) and has a better scalability in comparison with classic experimental design methods. Our in-depth analysis shows that the new method tends to favor experiments that are on the one side hard-to-predict and on the other side representative for the rest of the experiments. Efficient optimization of the new design problem is achieved through alternating optimization and sequential greedy search. Extensive experimental results on synthetic problems and three real-world tasks, including questionnaire design for preference learning, active learning for text categorization, and spatial sensor placement, highlight the advantages of the proposed approaches.

Ordinal regression has become an effective way of learning user preferences, but most research focuses on single regression problems. In this paper we introduce collaborative ordinal regression, where multiple ordinal regression tasks are handled simultaneously. Rather than modeling each task individually, we explore the dependency between ranking functions through a hierarchical Bayesian model and assign a common Gaussian Process (GP) prior to all individual functions. Empirical studies show that our collaborative model outperforms the individual counterpart in preference learning applications.

Principal component analysis (PCA) has been extensively applied in data mining, pattern recognition and information retrieval for unsupervised dimensionality reduction. When labels of data are available, e.g., in a classification or regression task, PCA is however not able to use this information. The problem is more interesting if only part of the input data are labeled, i.e., in a semi-supervised setting. In this paper we propose a supervised PCA model called SPPCA and a semi-supervised PCA model called S2PPCA, both of which are extensions of a probabilistic PCA model. The proposed models are able to incorporate the label information into the projection phase, and can naturally handle multiple outputs (i.e., in multi-task learning problems). We derive an efficient EM learning algorithm for both models, and also provide theoretical justifications of the model behaviors. SPPCA and S2PPCA are compared with other supervised projection methods on various learning tasks, and show not only promising performance but also good scalability.

We introduce a Gaussian process (GP) framework, stochastic relational models (SRM), for learning social, physical, and other relational phenomena where interactions between entities are observed. The key idea is to model the stochastic structure of entity relationships (i.e., links) via a tensor interaction of multiple GPs, each defined on one type of entities. These models in fact define a set of nonparametric priors on infinite dimensional tensor matrices, where each element represents a relationship between a tuple of entities. By maximizing the marginalized likelihood, information is exchanged between the participating GPs through the entire relational network, so that the dependency structure of links is messaged to the dependency of entities, reflected by the adapted GP kernels. The framework offers a discriminative approach to link prediction, namely, predicting the existences, strengths, or types of relationships based on the partially observed linkage network as well as the attributes of entities (if given). We discuss properties and variants of SRM and derive an efficient learning algorithm. Very encouraging experimental results are achieved on a toy problem and a user-movie preference link prediction task. In the end we discuss extensions of SRM to general relational learning tasks.

Dimensionality reduction by feature projection is widely used in pattern recognition, information retrieval, and statistics. When there are some outputs available (e.g., regression values or classification results), it is often beneficial to consider supervised projection, which is based not only on the inputs, but also on the target values. While this applies to a single-output setting, we are more interested in applications with multiple outputs, where several tasks need to be learned simultaneously. In this paper, we introduce a novel projection approach called Multi-Output Regularized feature Projection (MORP), which preserves the information of input features and, meanwhile, captures the correlations between inputs/outputs and (if applicable) between multiple outputs. This is done by introducing a latent variable model on the joint input-output space and minimizing the reconstruction errors for both inputs and outputs. It turns out that the mappings can be found by solving a generalized eigenvalue problem and are ready to extend to nonlinear mappings. Prediction accuracy can be greatly improved by using the new features since the structure of outputs is explored. We validate our approach in two applications. In the first setting, we predict users' preferences for a set of paintings. The second is concerned with image and text categorization where each image (or document) may belong to multiple categories. The proposed algorithm produces very encouraging results in both settings.

Dimensionality reduction via feature projection has been widely used in pattern recognition and machine learning. It is often beneficial to derive the projections not only based on the inputs but also on the target values in the training data set. This is of particular importance in predicting multivariate or structured outputs which is an area of growing interest. In this paper we introduce a novel projection framework which is sensitive to both input features and outputs. Based on the derived features prediction accuracy can be greatly improved. We validate our approach in two applications. The first is to model users' preferences on a set of paintings. The second application is concerned with image categorization where each image may belong to multiple categories. The proposed algorithm produces very encouraging results in both settings.

Organizing textual documents into a hierarchical taxonomy is a common practice in knowledge management. Beside textual features, the hierarchical structure of directories reflect additional and important knowledge annotated by experts. It is generally desired to incorporate this information into text mining processes. In this paper, we propose hierarchy-regularized latent semantic indexing, which encodes the hierarchy into a similarity graph of documents and then formulates an optimization problem mapping each document into a low dimensional vector space. The new feature space preserves the intrinsic structure of the original taxonomy and thus provides a meaningful basis for various learning tasks like visualization and classification. Our approach employs the information about class proximity and class specificity, and can naturally cope with multi-labeled documents. Our empirical studies show very encouraging results on two real-world data sets, the new Reuters (RCVI) benchmark and the Swissprot protein database.

We apply nonparametric hierarchical Bayesian modelling to relational learning. In a hierarchical Bayesian approach, model parameters can be "personalized", i.e., owned by entities or relationships, and are coupled via a common prior distribution. Flexibility is added in a nonparametric hierarchical Bayesian approach, such that the learned knowledge can be truthfully represented. We apply our approach to a medical domain where we form a nonparametric hierarchical Bayesian model for relations involving hospitals, patients, procedures and diagnosis. The experiments show that the additional flexibility in a nonparametric hierarchical Bayes approach results in a more accurate model of the dependencies between procedures and diagnosis and gives significantly improved estimates of the probabilities of future procedures.

We consider the problem of multi-task learning, that is, learning multiple related functions. Our approach is based on a hierarchical Bayesian framework, that exploits the equivalence between parametric linear models and nonparametric Gaussian processes (GPs). The resulting models can be learned easily via an EM-algorithm. Empirical studies on multi-label text categorization suggest that the presented models allow accurate solutions of these multi-task problems.

We propose a simple clustering framework on graphs encoding pairwise data similarities. Unlike usual similarity-based methods, the approach softly assigns data to clusters in a probabilistic way. More importantly, a hierarchical clustering is naturally derived in this framework to gradually merge lower-level clusters into higher-level ones. A random walk analysis indicates that the algorithm exposes clustering structures in various resolutions, i.e., a higher level statistically models a longer-term diffusion on graphs and thus discovers a more global clustering structure. Finally we provide very encouraging experimental results.

Latent semantic indexing (LSI) is a well-known unsupervised approach for dimensionality reduction in information retrieval. However if the output information (i.e. category labels) is available, it is often beneficial to derive the indexing not only based on the inputs but also on the target values in the training data set. This is of particular importance in applications with multiple labels, in which each document can belong to several categories simultaneously. In this paper we introduce the multi-label informed latent semantic indexing (MLSI) algorithm which preserves the information of inputs and meanwhile captures the correlations between the multiple outputs. The recovered "latent semantics" thus incorporate the human-annotated category information and can be used to greatly improve the prediction accuracy. Empirical study based on two data sets, Reuters-21578 and RCV1, demonstrates very encouraging results.

We present a novel method for learning with Gaussian process regres- sion in a hierarchical Bayesian framework. In a first step, kernel matri- ces on a fixed set of input points are learned from data using a simple and efficient EM algorithm. This step is nonparametric, in that it does not require a parametric form of covariance function. In a second step, kernel functions are fitted to approximate the learned covariance matrix using a generalized Nystrom method, which results in a complex, data driven kernel. We evaluate our approach as a recommendation engine for art images, where the proposed hierarchical Bayesian method leads to excellent prediction performance. 1 Introduction In many real-world application domains, the available training data sets are quite small, which makes learning and model selection difficult. For example, in the user preference modelling problem we will consider later, learning a preference model would amount to fitting a model based on only 20 samples of a user's preference data. Fortunately, there are situations where individual data sets are small, but data from similar scenarios can be obtained. Returning to the example of preference modelling, data for many different users are typically available. This data stems from clearly separate individuals, but we can expect that models can borrow strength from data of users with similar tastes. Typically, such problems have been handled by either mixed effects models or hierarchical Bayesian modelling. In this paper we present a novel approach to hierarchical Bayesian modelling in the context of Gaussian process regression, with an application to recommender systems. Here, hier- archical Bayesian modelling essentially means to learn the mean and covariance function of the Gaussian process. In a first step, a common collaborative kernel matrix is learned from the data via a simple and efficient EM algorithm. This circumvents the problem of kernel design, as no paramet- ric form of kernel function is required here. Thus, this form of learning a covariance matrix is also suited for problems with complex covariance structure (e.g. nonstationarity). A portion of the learned covariance matrix can be explained by the input features and, thus, generalized to new objects via a content-based kernel smoother. Thus, in a second step, we generalize the covariance matrix (learned by the EM-algorithm) to new items using a generalized Nystrom method. The result is a complex content-based kernel which itself is a weighted superposition of simple smoothing kernels. This second part could also be applied to other situations where one needs to extrapolate a covariance matrix on a finite set (e.g. a graph) to a continuous input space, as, for example, required in induction for semi-supervised learning [14]. The paper is organized as follows. Sec. 2 casts Gaussian process regression in a hierarchical Bayesian framework, and shows the EM updates to learn the covariance matrix in the first step. Extrapolating the covariance matrix is shown in Sec. 3. We illustrate the function of the EM-learning on a toy example in Sec. 4, before applying the proposed methods as a recommender system for images in Sec. 4.1. 1.1 Previous Work In statistics, modelling data from related scenarios is typically done via mixed effects mod- els or hierarchical Bayesian (HB) modelling [6]. In HB, parameters of models for indi- vidual scenarios (e.g. users in recommender systems) are assumed to be drawn from a common (hyper)prior distribution, allowing the individual models to interact and regular- ize each other. Recent examples of HB modelling in machine learning include [1, 2]. In other contexts, this learning framework is called multi-task learning [4]. Multi-task learn- ing with Gaussian processes has been suggested by [8], yet with the rather stringent as- sumption that one has observations on the same set of points in each individual scenario. Based on sparse approximations of GPs, a more general GP multi-task learner with para- metric covariance functions has been presented in [7]. In contrast, the approach presented in this paper only considers covariance matrices (and is thus non-parametric) in the first step. Only in a second extrapolation step, kernel smoothing leads to predictions based on a covariance function that is a data-driven combination of simple kernel functions. 2 Learning GP Kernel Matrices via EM The learning task we are concerned with can be stated as follows: The data are observations from M different scenarios. In the i.th scenario, we have observations yi = (yi , . . . , yi ) 1 N i on a total of N i points, Xi = {xi , . . . , xi } 1 . In order to analyze this data in a hierarchical N i Bayesian way, we assume that the data for each scenario is a noisy sample of a Gaussian process (GP) with unknown mean and covariance function. We assume that mean and covariance function are shared across different scenarios.1 In the first modelling step presented in this section, we consider transductive learning ("la- belling a partially labelled data set"), that is, we are interested in the model's behavior only on points X, with X = M Xi and cardinality N = |X|. This situation is relevant i=1 for most collaborative filtering applications. Thus, test points are the unlabelled points in each scenario. This reduces the whole "infinite dimensional" Gaussian process to its finite dimensional projection on points X, which is an N -variate Gaussian distribution with co- variance matrix K and mean vector m. For the EM algorithm to work, we also require that there is some overlap between scenarios, that is, Xi Xj = for some i, j. Coming back to the user modelling problem mentioned above, this means that at least some items have been rated by more than one user. Thus, our first modelling step focusses on directly learning the covariance matrix K and 1Alternative HB approaches for collaborative filtering, like that discussed in [5], assume that model weights are drawn from a shared Gaussian distribution. m from the data via an efficient EM algorithm. This may be of particular help in problems where one would need to specify a complex (e.g. nonstationary) covariance function. Following the hierarchical Bayesian assumption, the data observed in each scenario is thus a partial sample from N (y | m, K + 21), where 1 denotes the unit matrix. The joint model is simply M p(m, K) p(yi | f i)p(f i | m, K), (1) i=1 where p(m, K) denotes the prior distribution for mean and covariance. We assume a Gaus- sian likelihood p(yi | f i) with diagonal covariance matrix 21. 2.1 EM Learning For the above hierarchical Bayesian model, Eq. (1), the marginal likelihood becomes M p(m, K) p(yi | f i)p(f i | m, K) df i. (2) i=1 To obtain simple and stable solutions when estimating m and K from the data, we con- sider point estimates of the parameters m and K, based on a penalized likelihood approach with conjugate priors.2 The conjugate prior for mean m and covariance K of a multivari- ate Gaussian is the so-called Normal-Wishart distribution [6], which decomposes into the product of an inverse Wishart distribution for K and a Normal distribution for m, p(m, K) = N (m | , -1K)Wi-1(K|, U ). (3) That is, the prior for the Gram matrix K is given by an inverse Wishart distribution with scalar parameter > 1/2(N - 1) and U being a symmetric positive-definite matrix. Given the covariance matrix K, m is Gaussian distributed with mean and covariance -1K, where is a positive scalar. The parameters can be interpreted in terms of an equivalent data set for the mean (this data set has size A, with A = , and mean = ) and a data set for the covariance that has size B, with = (B + N )/2, and covariance S, U = (B/2)S. In order to write down the EM algorithm in a compact way, we denote by I(i) the set of indices of those data points that have been observed in the i.th scenario, that is I(i) = {j | j {1, . . . , N } and xj Xi}. Keep in mind that in most applications of interest N i N such that most targets are missing in training. KI(i),I(i) denotes the square submatrix of K that corresponds to points I(i), that is, the covariance matrix for points in the i.th scenario. By K,I(i) we denote the covariance matrix of all N points versus those in the i.th scenario.

Information filtering has made considerable progress in recent years. The predominant approaches are content-based methods and collaborative methods. Researchers have largely concentrated on either of the two approaches since a principled unifying framework is still lacking. This paper suggests that both approaches can be combined under a hierarchical Bayesian framework. Individual content-based user profiles are generated and collaboration between various user models is achieved via a common learned prior distribution. However, it turns out that a parametric distribution (e.g. Gaussian) is too restrictive to describe such a common learned prior distribution. We thus introduce a nonparametric common prior, which is a sample generated from a Dirichlet process which assumes the role of a hyper prior. We describe effective means to learn this nonparametric distribution, and apply it to learn users' information needs. The resultant algorithm is simple and understandable, and offers a principled solution to combine content-based filtering and collaborative filtering. Within our framework, we are now able to interpret various existing techniques from a unifying point of view. Finally we demonstrate the empirical success of the proposed information filtering methods.

Memory-based collaborative filtering (CF) has been studied extensively in the literature and has proven to be successful in various types of personalized recommender systems. In this paper, we develop a probabilistic framework for memory-based CF (PMCF). While this framework has clear links with classical memory-based CF, it allows us to find principled solutions to known problems of CF-based recommender systems. In particular, we show that a probabilistic active learning method can be used to actively query the user, thereby solving the "new user problem." Furthermore, the probabilistic framework allows us to reduce the computational cost of memory-based CF by working on a carefully selected subset of user profiles, while retaining high accuracy. We report experimental results based on two real-world data sets, which demonstrate that our proposed PMCF framework allows an accurate and efficient prediction of user preferences.

This paper aims to address the problem of art image retrieval (AIR), which aims to help users find their favorite painting images. AIR is of great interests to us because of its application potentials and interesting research challenges---the retrieval is not only based on painting contents or styles, but also heavily based on user preference profiles. This paper describes the collaborative ensemble learning, a novel statistical learning approach to this task. It at first applies probabilistic support vector machines (SVMs) to model each individual user's profile based on given examples, i.e. liked or disliked paintings. Due to the high complexity of profile modelling, the SVMs can be rather weak in predicting preferences for new paintings. To overcome this problem, we combine a society of users' profiles, represented by their respective SVM models, to predict a given user's preferences for painting images. We demonstrate that the combination scheme is embedded in a Bayesian framework and retains intuitive interpretations---like-minded users are likely to share similar preferences. We report extensive empirical studies based on two experimental settings. The first one includes some controlled simulations performed on 4533 painting images. In the second setting, we report evaluations based on user preferences collected through an online web-based survey. Both experiments demonstrate that the proposed approach achieves excellent performance in terms of capturing a user's diverse preferences.

The application range of memory-based collaborative filtering (CF) is limited due to CF's high memory consumption and long runtime. The approach presented in this paper removes redundant and inconsistent instances (users) from the data. This paper aims to distinguish informative instances (users) from large raw user preference database and thus alleviate the memory and runtime cost of the widely used memory-based collaborative filtering (CF) algorithm. Our work shows that a satisfactory accuracy can be achieved by using only a small portion of the original data set, thereby alleviating the storage and runtime cost of the CF algorithm. In our approach, we consider instance selection as the problem of selecting informative data that increase the We begin by discussing the instance selection problem in a general sense that is to increase a posteriori probability of the optimal model by selecting informative data. We evaluate the empirical performance of our approach PF on two real-world data sets and attain very promisingpositive experimental results. The dData size and the prediction time are significantly reduced, while the prediction accuracy is on a par with almost the same as the results achieved by using the complete database.

Collaborative filtering (CF) has become an important data mining technique to make personalized recommendations for books, web pages or movies, etc. One popular algorithm is the memory-based collaborative filtering, which predicts a user's preference based on his or her similarity to other users (instances) in the database. However, the tremendous growth of users and the large number of products, memory-based CF algorithms results in the problem of deciding the right instances to use during prediction, in order to reduce executive cost and excessive storage, and possibly to improve the generalization accuracy by avoiding noise and overfitting. In this paper, we focus our work on a typical user preference database that contains many missing values, and propose four novel instance reduction techniques called TURF1-TURF4 as a preprocessing step to improve the efficiency and accuracy of the memory-based CF algorithm. The key idea is to generate prediction from a carefully selected set of relevant instances. We evaluate the techniques on the well-known EachMovie data set. Our experiments showed that the proposed algorithms not just dramatically speed up the prediction, but also improved the accuracy.

Collaborative filtering uses a database about consumers' preferences to make personal product recommendations and is achieving widespread success in both E-Commerce and Information Filtering Applications nowadays. However, the traditional collaborative filtering algorithms do not scale well to the ever-growing number of consumers. The quality of the recommendation also needs to be improved in order to gain more trust from the consumers. In this paper, we present a novel method to improve the scalability and the accuracy of the collaborative filtering algorithm. We introduce an information theoretic approach to measure the relevance of a consumer (instance) for predicting the preference for the given product (target concept). The proposed method reduces the training data set by selecting only highly relevant instances. Our experimental evaluation on the well-known EachMovie data set shows that our method doesn't only significantly speed up the prediction, but also results in a better accuracy.