For the past several years, we have been using robots in our introductory computer science course. Although this has been challenging for many reasons, it has also been very rewarding on a number of fronts, both for the students and for us. However, in order for this to occur, we had to adapt to what we perceived as “chaotic code.” In this paper we describe lessons learned by watching what the students do, where they have trouble, and what they enjoy. Further, we discuss what the implications of focusing on creativity has had on teaching and assessment.

In many AI domains such as product configuration, a user should interactively specify a solution that must satisfy a set of constraints. In such scenarios, offline compilation of feasible solutions into a tractable representation is an important approach to delivering efficient backtrack-free user interaction online. In particular,binary decision diagrams (BDDs) have been successfully used as a compilation target for product and service configuration. In this paper we discuss how to extend BDD-based configuration to scenarios involving cost functions which express user preferences. We first show that an efficient, robust and easy to implement extension is possible if the cost function is additive, and feasible solutions are represented using multi-valued decision diagrams (MDDs). We also discuss the effect on MDD size if the cost function is non-additive or if it is encoded explicitly into MDD. We then discuss interactive configuration in the presence of multiple cost functions. We prove that even in its simplest form, multiple-cost configuration is NP-hard in the input MDD. However, for solving two-cost configuration we develop a pseudo-polynomial scheme and a fully polynomial approximation scheme. The applicability of our approach is demonstrated through experiments over real-world configuration models and product-catalogue datasets. Response times are generally within a fraction of a second even for very large instances.

It has been established that the second-order stochastic gradient descent (2SGD) method can potentially achieve generalization performance as well as empirical optimum in a single pass (i.e., epoch) through the training examples. However, 2SGD requires computing the inverse of the Hessian matrix of the loss function, which is prohibitively expensive. This paper presents Periodic Step-size Adaptation (PSA), which approximates the Jacobian matrix of the mapping function and explores a linear relation between the Jacobian and Hessian to approximate the Hessian periodically and achieve near-optimal results in experiments on a wide variety of models and tasks.

We propose a novel information theoretic approach for semi-supervised learning of conditional random fields. Our approach defines a training objective that combines the conditional likelihood on labeled data and the mutual information on unlabeled data. Different from previous minimum conditional entropy semi-supervised discriminative learning methods, our approach can be naturally cast into the rate distortion theory framework in information theory. We analyze the tractability of the framework for structured prediction and present a convergent variational training algorithm to defy the combinatorial explosion of terms in the sum over label configurations. Our experimental results show that the rate distortion approach outperforms standard $l_2$ regularization and minimum conditional entropy regularization on both multi-class classification and sequence labeling problems.

The stochastic approximation method is behind the solution to many important, actively-studied problems in machine learning. Despite its far-reaching application, there is almost no work on applying stochastic approximation to learning problems with constraints. The reason for this, we hypothesize, is that no robust, widely-applicable stochastic approximation method exists for handling such problems. We propose that interior-point methods are a natural solution. We establish the stability of a stochastic interior-point approximation method both analytically and empirically, and demonstrate its utility by deriving an on-line learning algorithm that also performs feature selection via L1 regularization.

Introduction In the last years, many useful NLP tools have been developed and many of them are now even available commercially. Most of these tools are monolingual or multi-monolingual, meaning that the software can deal with more than one language, but that the re sults will al ways be displayed in the same language as the text. We therefore distinguish these applications from cross-lingual software, which is software that helps to transgress the language bound ary. Examples for such applications are machine translation and cross-lingual document retrieval, i.e. retrieval using search engines which allow to en ter a search term in one language and which also yield results in other languages, usually because the query is translated in one way or another. In our eyes, cross-lingual applications are currently the bottleneck of available NLP tools. To our knowledge, there are no applications that allow comparing documents written in dif ferent languages with each other and there are very few which give users a quick overview of the ap proximate contents of documents written in different languages.

This paper presents the design of an associative memory with feedback that is capable of on-line temporal sequence learning. A framework for on-line sequence learning has been proposed, and different sequence learning models have been analysed according to this framework. The network model is an associative memory with a separate store for the sequence context of a symbol. A sparse distributed memory is used to gain scalability. The context store combines the functionality of a neural layer with a shift register. The sensitivity of the machine to the sequence context is controllable, resulting in different characteristic behaviours. The model can store and predict on-line sequences of various types and length. Numerical simulations on the model have been carried out to determine its properties.

This paper belongs to the area of learning theory that has been variously referred to as prediction with expert advice, competitive on-line prediction, p rediction of individual sequences, and universal on-line learning; see [7] for a re view. There are many proof techniques known in this field; this paper is based on K alnishkan and Vyugin's Weak Aggregating Algorithm [16], but it is possible that som e of the numerous other techniques could be used instead. In Section 2 we give the main definitions and state our main results, Th e-orems 1-4; their proofs are given in Sections 3-6. In Section 7 we inf ormally discuss the notion of stationarity, and Section 8 concludes.

The evolution in health care to greater support for self-managed care is escalating the demand for e-health systems in which patients can access their personal health information in order to ultimately partner with providers in the management of their health and wellness care. At present, unfortunately, patients are seldom able to easily access their own health information so, as a result, it is often difficult for patients to enter into a dialogue with their healthcare providers about treatment and other options. One truism seems to be constantly ignored: it is not possible for patients to actively manage their health without the requisite information. Health information should be made available through "any time, anywhere" delivery: outside the physician's office or hospital, in the home or other personal setting, on a variety of multimedia information devices. We believe that personalization of health information will be a key element in effective self-managed healthcare.

The goal of this paper is to present four key assumptions regarding the measurement of cognitive and metacognitive regulatory processes used during learning with hypermedia. First, we assume it is possible to detect, trace, model, and foster SRL processes during learning with hypermedia. Second, understanding the complex nature of the regulatory processes during learning with hypermedia is critical in determining why certain processes are used throughout a learning task. Third, it is assumed that the use of SRL processes can dynamically change over time and that they are cyclical in nature (influenced by internal and external conditions and feedback mechanisms). Fourth, capturing, identifying, and classifying SRL processes used during learning with hypermedia is a rather challenging task.