Graph representation learning has achieved a remarkable success in many graph-based applications, such as node classification, link prediction, and community detection. These models are usually designed to preserve the vertex information at different granularity and reduce the problems in discrete space to some machine learning tasks in continuous space. However, regardless of the fruitful progress, for some kind of graph applications, such as graph compression and edge partition, it is very hard to reduce them to some graph representation learning tasks. Moreover, these problems are closely related to reformulating a global layout for a specific graph, which is an important NP-hard combinatorial optimization problem: graph ordering. In this paper, we propose to attack the graph ordering problem behind such applications by a novel learning approach. Distinguished from greedy algorithms based on predefined heuristics, we propose a neural network model: Deep Order Network (DON) to capture the hidden locality structure from partial vertex order sets. Supervised by sampled partial order, DON has the ability to infer unseen combinations. Furthermore, to alleviate the combinatorial explosion in the training space of DON and make the efficient partial vertex order sampling , we employ a reinforcement learning model: the Policy Network, to adjust the partial order sampling probabilities during the training phase of DON automatically. To this end, the Policy Network can improve the training efficiency and guide DON to evolve towards a more effective model automatically. Comprehensive experiments on both synthetic and real data validate that DON-RL outperforms the current state-of-the-art heuristic algorithm consistently. Two case studies on graph compression and edge partitioning demonstrate the potential power of DON-RL in real applications.

Machine learning models, used to predict future use of primary care services from the Veterans Affairs (VA) Health Care System, did not outperform traditional regression models. ABSTRACT Objectives: The Veterans Affairs (VA) Health Care System is among the largest integrated health systems in the United States. Many VA enrollees are dual users of Medicare, and little research has examined methods to most accurately predict which veterans will be mostly reliant on VA services in the future. This study examined whether machine learning methods can better predict future reliance on VA primary care compared with traditional statistical methods. Study Design: Observational study of 83,143 VA patients dually enrolled in fee-for-service Medicare using VA and Medicare administrative databases and the 2012 Survey of Healthcare Experiences of Patients.

The output shows that the values of the various variables are not standardized. For example, the V14 variables, which is our target variable, it holds only 2 values, either 0 or 1. Instead, this should be a categorical variable.

The Mixed-Membership Stochastic Blockmodel~(MMSB) is proposed as one of the state-of-the-art Bayesian relational methods suitable for learning the complex hidden structure underlying the network data. However, the current formulation of MMSB suffers from the following two issues: (1), the prior information~(e.g. entities' community structural information) can not be well embedded in the modelling; (2), community evolution can not be well described in the literature. Therefore, we propose a non-parametric fragmentation coagulation based Mixed Membership Stochastic Blockmodel (fcMMSB). Our model performs entity-based clustering to capture the community information for entities and linkage-based clustering to derive the group information for links simultaneously. Besides, the proposed model infers the network structure and models community evolution, manifested by appearances and disappearances of communities, using the discrete fragmentation coagulation process (DFCP). By integrating the community structure with the group compatibility matrix we derive a generalized version of MMSB. An efficient Gibbs sampling scheme with Polya Gamma (PG) approach is implemented for posterior inference. We validate our model on synthetic and real world data.

The emerge of new technologies to synthesize and analyze big data with high-performance computing, has increased our capacity to more accurately predict crop yields. Recent research has shown that Machine learning (ML) can provide reasonable predictions, faster, and with higher flexibility compared to simulation crop modeling. The earlier the prediction during the growing season the better, but this has not been thoroughly investigated as previous studies considered all data available to predict yields. This paper provides a machine learning based framework to forecast corn yields in three US Corn Belt states (Illinois, Indiana, and Iowa) considering complete and partial in-season weather knowledge. Several ensemble models are designed using blocked sequential procedure to generate out-of-bag predictions. The forecasts are made in county-level scale and aggregated for agricultural district, and state level scales. Results show that ensemble models based on weighted average of the base learners outperform individual models. Specifically, the proposed ensemble model could achieve best prediction accuracy (RRMSE of 7.8%) and least mean bias error (-6.06 bu/acre) compared to other developed models. Comparing our proposed model forecasts with the literature demonstrates the superiority of forecasts made by our proposed ensemble model. Results from the scenario of having partial in-season weather knowledge reveal that decent yield forecasts can be made as early as June 1st. To find the marginal effect of each input feature on the forecasts made by the proposed ensemble model, a methodology is suggested that is the basis for finding feature importance for the ensemble model. The findings suggest that weather features corresponding to weather in weeks 18-24 (May 1st to June 1st) are the most important input features.

In programmatic advertising, ad slots are usually sold using second-price (SP) auctions in real-time. The highest bidding advertiser wins but pays only the second-highest bid (known as the winning price). In SP, for a single item, the dominant strategy of each bidder is to bid the true value from the bidder's perspective. However, in a practical setting, with budget constraints, bidding the true value is a sub-optimal strategy. Hence, to devise an optimal bidding strategy, it is of utmost importance to learn the winning price distribution accurately. Moreover, a demand-side platform (DSP), which bids on behalf of advertisers, observes the winning price if it wins the auction. For losing auctions, DSPs can only treat its bidding price as the lower bound for the unknown winning price. In literature, typically censored regression is used to model such partially observed data. A common assumption in censored regression is that the winning price is drawn from a fixed variance (homoscedastic) uni-modal distribution (most often Gaussian). However, in reality, these assumptions are often violated. We relax these assumptions and propose a heteroscedastic fully parametric censored regression approach, as well as a mixture density censored network. Our approach not only generalizes censored regression but also provides flexibility to model arbitrarily distributed real-world data. Experimental evaluation on the publicly available dataset for winning price estimation demonstrates the effectiveness of our method. Furthermore, we evaluate our algorithm on one of the largest demand-side platforms and significant improvement has been achieved in comparison with the baseline solutions.

We investigate the framework of privacy amplification by iteration, recently proposed by Feldman et al., from an information-theoretic lens. We demonstrate that differential privacy guarantees of iterative mappings can be determined by a direct application of contraction coefficients derived from strong data processing inequalities for f -divergences. In particular, by generalizing the Dobrushin's contraction coefficient for total variation distance to an f -divergence known as E Differential privacy (DP) [1, 2] has become the standard definition for designing privacy-preserving machine learning algorithms. One reason for its success is its operational significance, which can be best described in terms of binary hypothesis testing (see, e.g., [3, 4]). Nevertheless, it is often difficult to compute DP guarantees for applications where a high number of data accesses is needed for a single analysis [5, 6]. To obtain the DP parameters in such applications, which include machine learning models trained using stochastic gradient descent (SGD), one needs to resort to composition theorems which are often loose due to their generality. As a remedy, several variants of DP have been recently proposed [7-10] based on Rényi divergence. These variants enjoy better composition properties.

Dimensionality reduction is considered as an important step for ensuring competitive performance in unsupervised learning such as anomaly detection. Non-negative matrix factorization (NMF) is a popular and widely used method to accomplish this goal. But NMF, together with its recent, enhanced version, like graph regularized NMF or symmetric NMF, do not have the provision to include the neighborhood structure information and, as a result, may fail to provide satisfactory performance in presence of nonlinear manifold structure. To address that shortcoming, we propose to consider and incorporate the neighborhood structural similarity information within the NMF framework by modeling the data through a minimum spanning tree. What motivates our choice is the understanding that in the presence of complicated data structure, a minimum spanning tree can approximate the intrinsic distance between two data points better than a simple Euclidean distance does, and consequently, it constitutes a more reasonable basis for differentiating anomalies from the normal class data. We label the resulting method as the neighborhood structure assisted NMF. By comparing the formulation and properties of the neighborhood structure assisted NMF with other versions of NMF including graph regularized NMF and symmetric NMF, it is apparent that the inclusion of the neighborhood structure information using minimum spanning tree makes a key difference. We further devise both offline and online algorithmic versions of the proposed method. Empirical comparisons using twenty benchmark datasets as well as an industrial dataset extracted from a hydropower plant demonstrate the superiority of the neighborhood structure assisted NMF and support our claim of merit.

There is a large body of work on convergence rates either in passive or active learning. Here we first outline some of the main results that have been obtained, more specifically in a nonparametric setting under assumptions about the smoothness of the regression function (or the boundary between classes) and the margin noise. We discuss the relative merits of these underlying assumptions by putting active learning in perspective with recent work on passive learning. We design an active learning algorithm with a rate of convergence better than in passive learning, using a particular smoothness assumption customized for k-nearest neighbors. Unlike previous active learning algorithms, we use a smoothness assumption that provides a dependence on the marginal distribution of the instance space. Additionally, our algorithm avoids the strong density assumption that supposes the existence of the density function of the marginal distribution of the instance space and is therefore more generally applicable.

We consider gradient descent with `momentum', a widely used method for loss function minimization in machine learning. This method is often used with `Nesterov acceleration', meaning that the gradient is evaluated not at the current position in parameter space, but at the estimated position after one step. In this work, we show that the algorithm can be improved by extending this `acceleration' --- by using the gradient at an estimated position several steps ahead rather than just one step ahead. How far one looks ahead in this `super-acceleration' algorithm is determined by a new hyperparameter. Considering a one-parameter quadratic loss function, the optimal value of the super-acceleration can be exactly calculated and analytically estimated. We show explicitly that super-accelerating the momentum algorithm is beneficial, not only for this idealized problem, but also for several synthetic loss landscapes and for the MNIST classification task with neural networks. Super-acceleration is also easy to incorporate into adaptive algorithms like RMSProp or Adam, and is shown to improve these algorithms.