We present a novel method for frequentist statistical inference in M-estimation problems, based on stochastic gradient descent (SGD) with a fixed step size: we demonstrate that the average of such SGD sequences can be used for statistical inference, after proper scaling. An intuitive analysis using the Ornstein-Uhlenbeck process suggests that such averages are asymptotically normal. To show the merits of our scheme, we apply it to both synthetic and real data sets, and demonstrate that its accuracy is comparable to classical statistical methods, while requiring potentially far less computation.

We propose semi-random features for nonlinear function approximation. The flexibility of semi-random feature lies between the fully adjustable units in deep learning and the random features used in kernel methods. For one hidden layer models with semi-random features, we prove with no unrealistic assumptions that the model classes contain an arbitrarily good function as the width increases (universality), and despite non-convexity, we can find such a good function (optimization theory) that generalizes to unseen new data (generalization bound). For deep models, with no unrealistic assumptions, we prove universal approximation ability, a lower bound on approximation error, a partial optimization guarantee, and a generalization bound. Depending on the problems, the generalization bound of deep semi-random features can be exponentially better than the known bounds of deep ReLU nets; our generalization error bound can be independent of the depth, the number of trainable weights as well as the input dimensionality. In experiments, we show that semi-random features can match the performance of neural networks by using slightly more units, and it outperforms random features by using significantly fewer units. Moreover, we introduce a new implicit ensemble method by using semi-random features.

We propose a framework for modeling and estimating the state of controlled dynamical systems, where an agent can affect the system through actions and receives partial observations. Based on this framework, we propose Predictive State Representation with Random Fourier Features (RFF-PSR). A key property in RFF-PSRs is that the state estimate is represented by a conditional distribution of future observations given future actions. RFFPSRs combine this representation with moment-matching, kernel embedding, and local optimization to achieve a method that enjoys several favorable qualities: It can represent controlled environments which can be affected by actions, it has an efficient and theoretically justified learning algorithm, it uses a non-parametric representation that has expressive power to represent continuous non-linear dynamics. We provide a detailed formulation, a theoretical analysis and an experimental evaluation that demonstrates the effectiveness of our method.

We propose a novel approach for using unsupervised boosting to create an ensemble of generative models, where models are trained in sequence to correct earlier mistakes. Our meta-algorithmic framework can leverage any existing base learner that permits likelihood evaluation, including recent deep expressive models. Further, our approach allows the ensemble to include discriminative models trained to distinguish real data from model-generated data. We show theoretical conditions under which incorporating a new model in the ensemble will improve the fit and empirically demonstrate the effectiveness of our black-box boosting algorithms on density estimation, classification, and sample generation on benchmark datasets for a wide range of generative models.

Adversarial learning of probabilistic models has recently emerged as a promising alternative to maximum likelihood. Implicit models such as generative adversarial networks (GAN) often generate better samples compared to explicit models trained by maximum likelihood. Yet, GANs sidestep the characterization of an explicit density which makes quantitative evaluations challenging. To bridge this gap, we propose Flow-GANs, a generative adversarial network for which we can perform exact likelihood evaluation, thus supporting both adversarial and maximum likelihood training. When trained adversarially, Flow-GANs generate high-quality samples but attain extremely poor log-likelihood scores, inferior even to a mixture model memorizing the training data; the opposite is true when trained by maximum likelihood. Results on MNIST and CIFAR-10 demonstrate that hybrid training can attain high held-out likelihoods while retaining visual fidelity in the generated samples.

Modelling sparse and large data sets is highly in demand yet challenging in recommender systems. With the computation only on the non-zero ratings, Poisson Factorization (PF) enabled by variational inference has shown its high efficiency in scalable recommendation, e.g., modeling millions of ratings. However, as PF learns the ratings by individual users on items with the Gamma distribution, it cannot capture the coupling relations between users (items) and the rating popularity (i.e., favorable rating scores that are given to one item) and rating sparsity (i.e., those users (items) with many zero ratings) for one item (user). This work proposes Coupled Poisson Factorization (CPF) to learn the couplings between users (items), and the user/item attributes (i.e., metadata) are integrated into CPF to form the Metadata-integrated CPF (mCPF) to not only handle sparse but also popular ratings in very large-scale data. Our empirical results show that the proposed models significantly outperform PF and address the key limitations in PF for scalable recommendation.

Parameter tying is a regularization method in which parameters (weights) of a machine learning model are partitioned into groups by leveraging prior knowledge and all parameters in each group are constrained to take the same value. In this paper, we consider the problem of parameter learning in Markov networks and propose a novel approach called automatic parameter tying (APT) that uses automatic instead of a priori and soft instead of hard parameter tying as a regularization method to alleviate overfitting. The key idea behind APT is to set up the learning problem as the task of finding parameters and groupings of parameters such that the likelihood plus a regularization term is maximized. The regularization term penalizes models where parameter values deviate from their group mean parameter value. We propose and use a block coordinate ascent algorithm to solve the optimization task. We analyze the sample complexity of our new learning algorithm and show that it yields optimal parameters with high probability when the groups are well separated. Experimentally, we show that our method improves upon L 2 regularization and suggest several pragmatic techniques for good practical performance.

Cell nuclei detection and fine-grained classification have been fundamental yet challenging problems in histopathology image analysis. Due to the nuclei tiny size, significant inter-/intra-class variances, as well as the inferior image quality, previous automated methods would easily suffer from limited accuracy and robustness. In the meanwhile, existing approaches usually deal with these two tasks independently, which would neglect the close relatedness of them. In this paper, we present a novel method of sibling fully convolutional network with prior objectness interaction (called SFCN-OPI) to tackle the two tasks simultaneously and interactively using a unified end-to-end framework. Specifically, the sibling FCN branches share features in earlier layers while holding respective higher layers for specific tasks. More importantly, the detection branch outputs the objectness prior which dynamically interacts with the fine-grained classification sibling branch during the training and testing processes. With this mechanism, the fine-grained classification successfully focuses on regions with high confidence of nuclei existence and outputs the conditional probability, which in turn benefits the detection through back propagation. Extensive experiments on colon cancer histology images have validated the effectiveness of our proposed SFCN-OPI and our method has outperformed the state-of-the-art methods by a large margin.

Understanding politics is challenging because the politics take the influence from everything. Even we limit ourselves to the political context in the legislative processes; we need a better understanding of latent factors, such as legislators, bills, their ideal points, and their relations. From the modeling perspective, this is difficult 1) because these observations lie in a high dimension that requires learning on low dimensional representations, and 2) because these observations require complex probabilistic modeling with latent variables to reflect the causalities. This paper presents a new model to reflect and understand this political setting, NIPEN, including factors mentioned above in the legislation. We propose two versions of NIPEN: one is a hybrid model of deep learning and probabilistic graphical model, and the other model is a neural tensor model. Our result indicates that NIPEN successfully learns the manifold of the legislative bill's text, and NIPEN utilizes the learned low-dimensional latent variables to increase the prediction performance of legislators' votings. Additionally, by virtue of being a domain-rich probabilistic model, NIPEN shows the hidden strength of the legislators' trust network and their various characteristics on casting votes.

Collaborative filtering is an important technique for recommendation. Whereas it has been repeatedly shown to be effective in previous work,its performance remains unsatisfactory in many real-world applications, especially those where the items or users are highly diverse. In this paper, we explore an ensemble-based framework to enhance thecapability of a recommender in handling diverse data. Specifically, we formulate a probabilistic model which integrates the items, the users, as well as the associations between them into a generative process. On top of this formulation, we further derive a progressive algorithm to construct an ensemble of collaborative filters. In each iteration, a new filter is derived from re-weighted entries and incorporated into the ensemble. It is noteworthy that while the algorithmic procedure of our algorithm is apparently similar to boosting, it is derived from an essentially different formulation and thus differs in several key technical aspects. We tested the proposed method on three large datasets, and observed substantial improvement over the state of the art, including L 2 Boost, an effective method based on boosting.