Specifying and implementing flexible human-computer dialogs, such as those used in kiosks and smart phone apps, is challenging because of the numerous and varied directions in which each user might steer a dialog. The objective of this research is to improve dialog specification and implementation. To do so we enriched a notation based on concepts from programming languages, especially partial evaluation, for specifying a variety of unsolicited reporting, mixed-initiative dialogs in a concise representation that serves as a design for dialog implementation. We also built a dialog mining system that extracts a specification in this notation from requirements. To demonstrate that such a specification provides a design for dialog implementation, we built a system that automatically generates an implementation of the dialog, called a stager, from it. These two components constitute a dialog modeling toolkit that automates dialog specification and implementation. These results provide a proof of concept and demonstrate the study of dialog specification and implementation from a programming languages perspective. The ubiquity of dialogs in domains such as travel, education, and health care combined with the demand for smart phone apps provide a landscape for further investigation of these results.

Bounded rationality, that is, decision-making and planning under resource limitations, is widely regarded as an important open problem in artificial intelligence, reinforcement learning, computational neuroscience and economics. This paper offers a consolidated presentation of a theory of bounded rationality based on information-theoretic ideas. We provide a conceptual justification for using the free energy functional as the objective function for characterizing bounded-rational decisions. This functional possesses three crucial properties: it controls the size of the solution space; it has Monte Carlo planners that are exact, yet bypass the need for exhaustive search; and it captures model uncertainty arising from lack of evidence or from interacting with other agents having unknown intentions. We discuss the single-step decision-making case, and show how to extend it to sequential decisions using equivalence transformations. This extension yields a very general class of decision problems that encompass classical decision rules (e.g.

ABSTRACT In this paper we present a new model and an algorithm for unsupervised clustering of 2-D data such as images. We assume that the data comes from a union of multilinear subspaces (UOMS) model, which is a specific structured case of the much studied union of subspaces (UOS) model. For segmentation under this model, we develop Multilinear Subspace Clustering (MSC) algorithm and evaluate its performance on the YaleB and Olivietti image data sets. We show that MSC is highly competitive with existing algorithms employing the UOS model in terms of clustering performance while enjoying improvement in computational complexity. Index Terms - subspace clustering, multilinear algebra, spectral clustering 1. INTRODUCTION Most clustering algorithms seek to detect disjoint clouds of data.

Ordinal data are often seen in real applications. Regular multicategory classification methods are not designed for this data type and a more proper treatment is needed. We consider a framework of ordinal classification which pools the results from binary classifiers together. An inherent difficulty of this framework is that the class prediction can be ambiguous due to boundary crossing. To fix this issue, we propose a noncrossing ordinal classification method which materializes the framework by imposing noncrossing constraints. An asymptotic study of the proposed method is conducted. We show by simulated and data examples that the proposed method can improve the classification performance for ordinal data without the ambiguity caused by boundary crossings.

University of South Carolina, Department of Mechanical Engineering, Nuclear Engineering Program, Columbia, SC 29201 Send proofs to: Anthony M. Scopatz scopatz@cec.sc.edu 541 Main Street, Columbia, SC 29208 Number of Pages: 35 Number of Tables: 0 Number of Figures: 11 Keywords: nuclear fuel cycle, gaussian process, dynamic time warping Abstract A method for quickly determining deployment schedules that meet a given fuel cycle demand is presented here. This algorithm is fast enough to perform in situ within low-fidelity fuel cycle simulators. It uses Gaussian process regression models to predict the production curve as a function of time and the number of deployed facilities. Each of these predictions is measured against the demand curve using the dynamic time warping distance. The minimum distance deployment schedule is evaluated in a full fuel cycle simulation, whose generated production curve then informs the model on the next optimization iteration. The method converges within five to ten iterations to a distance that is less than one percent of the total deployable production. A representative once-through fuel cycle is used to demonstrate the methodology for reactor deployment. I INTRODUCTION With the recent advent of agent-based nuclear fuel cycle simulators, such as Cyclus [1, 2], there comes the possibility to make in situ, dynamic facility deployment decisions. This would more fully model real-world fuel cycles where institutions (such as utility companies) predict future demand and choose their future deployment schedules appropriately. However, one of the major challenges to making in situ deployment decisions is the speed at which "good enough" decisions can be made. This paper proposes three related deployment-specific optimization algorithms that can be used for any demand curve and facility type. The demands of a fuel cycle scenario can often be simply stated, e.g. Here, the dynamic time warping (DTW) [3] distance is minimized between the demand curve and the regression of a Gaussian Process model (GP) [4] of prior simulations. This minimization produces a guess for a deployment schedule which is subsequently tested using an actual simulator. This process is repeated until an optimal deployment schedule for the given demand is found. Importantly, by using the Gaussian process surrogates, the number of simulation realizations that must be executed as part of the optimization may be reduced to only a handful. Furthermore, it is at least two orders-of-magnitude faster to test the model than it is to run a single low-fidelity fuel cycle simulation. Because of the relative computational cheapness, it is suitable to be used inside of a fuel cycle simulation.

Motivated by vision-based reinforcement learning (RL) problems, in particular Atari games from the recent benchmark Aracade Learning Environment (ALE), we consider spatio-temporal prediction problems where future (image-)frames are dependent on control variables or actions as well as previous frames. While not composed of natural scenes, frames in Atari games are high-dimensional in size, can involve tens of objects with one or more objects being controlled by the actions directly and many other objects being influenced indirectly, can involve entry and departure of objects, and can involve deep partial observability. We propose and evaluate two deep neural network architectures that consist of encoding, action-conditional transformation, and decoding layers based on convolutional neural networks and recurrent neural networks. Experimental results show that the proposed architectures are able to generate visually-realistic frames that are also useful for control over approximately 100-step action-conditional futures in some games. To the best of our knowledge, this paper is the first to make and evaluate long-term predictions on high-dimensional video conditioned by control inputs.

We study modeling and inference with the Elliptical Gamma Distribution (EGD). We consider maximum likelihood (ML) estimation for EGD scatter matrices, a task for which we develop new fixed-point algorithms. Our algorithms are efficient and converge to global optima despite nonconvexity. Moreover, they turn out to be much faster than both a well-known iterative algorithm of Kent & Tyler (1991) and sophisticated manifold optimization algorithms. Subsequently, we invoke our ML algorithms as subroutines for estimating parameters of a mixture of EGDs. We illustrate our methods by applying them to model natural image statistics---the proposed EGD mixture model yields the most parsimonious model among several competing approaches.

We investigate a local reparameterizaton technique for greatly reducing the variance of stochastic gradients for variational Bayesian inference (SGVB) of a posterior over model parameters, while retaining parallelizability. This local reparameterization translates uncertainty about global parameters into local noise that is independent across datapoints in the minibatch. Such parameterizations can be trivially parallelized and have variance that is inversely proportional to the minibatch size, generally leading to much faster convergence. Additionally, we explore a connection with dropout: Gaussian dropout objectives correspond to SGVB with local reparameterization, a scale-invariant prior and proportionally fixed posterior variance. Our method allows inference of more flexibly parameterized posteriors; specifically, we propose variational dropout, a generalization of Gaussian dropout where the dropout rates are learned, often leading to better models. The method is demonstrated through several experiments.

The covariance matrix of a $p$-dimensional random variable is a fundamental quantity in data analysis. Given $n$ i.i.d. observations, it is typically estimated by the sample covariance matrix, at a computational cost of $O(np^{2})$ operations. When $n,p$ are large, this computation may be prohibitively slow. Moreover, in several contemporary applications, the population matrix is approximately sparse, and only its few large entries are of interest. This raises the following question, at the focus of our work: Assuming approximate sparsity of the covariance matrix, can its large entries be detected much faster, say in sub-quadratic time, without explicitly computing all its $p^{2}$ entries? In this paper, we present and theoretically analyze two randomized algorithms that detect the large entries of an approximately sparse sample covariance matrix using only $O(np\text{ poly log } p)$ operations. Furthermore, assuming sparsity of the population matrix, we derive sufficient conditions on the underlying random variable and on the number of samples $n$, for the sample covariance matrix to satisfy our approximate sparsity requirements. Finally, we illustrate the performance of our algorithms via several simulations.

Abstract--We use machine learning for designing a medium frequency trading strategy for a portfolio of 5 year and 10 year US Treasury note futures. We formulate this as a classification problem where we predict the weekly direction of movement of the portfolio using features extracted from a deep belief network trained on technical indicators of the portfolio constituents. The experimentation shows that the resulting pipeline is effective in making a profitable trade. I. INTRODUCTION AND RELATED WORK Machine learning application in finance is a challenging problem owing to low signal to noise ratio. Moreover, domain expertise is essential for engineering features which assist in solving an appropriate classification or regression problem.