The number of people with disabilities is continuously increasing. Providing patients who have disabilities with the rehabilitation and care necessary to allow them good quality of life creates overwhelming demands for health and rehabilitation services. We suggest that advancements in intelligent agent technology provide new opportunities for improving the provided services. We will discuss the challenges of building an agent for the health care domain and present four capabilities that are required for an agent in the health care domain: planning, monitoring, intervention and encouragement. We will discuss the importance of personalizing all of them and the needto facilitate cooperation between the automated agent and the human care givers. We will review recent technology that can be used toward the development of agents that can have these capabilities and their promise in automating services such as physiotherapy, speech therapy and cognitive training.

The selection of kernel function which determines the mapping between the input space and the feature space is of crucial importance to kernel methods. Existing kernel selection approaches commonly use some measures of generalization error, which are usually difficult to estimate and have slow convergence rates. In this paper, we propose a novel measure, called eigenvalues ratio (ER), of the tight bound of generalization error for kernel selection. ER is the ration between the sum of the main eigenvalues and that of the tail eigenvalues of the kernel matrix. Defferent from most of existing measures, ER is defined on the kernel matrxi, so it can be estimated easily from the available training data, which makes it usable for kernel selection. We establish tight ER-based generalization error bounds of order $O(\frac{1}{n})$ for several kernel-based methods under certain general conditions, while for most of existing measures, the convergence rate is at most $O(\frac{1}{\sqrt{n}})$. Finally, to guarantee good generalization performance, we propose a novel kernel selection criterion by minimizing the derived tight generalization error bounds. Theoretical analysis and experimental results demonstrate that our kernel selection criterion is a good choice for kernel seletion.

PDDL+ planning involves reasoning about mixed discrete-continuous change over time. Nearly all PDDL+ planners assume that continuous change is linear. We present a new technique that accommodates nonlinear change by encoding problems as nonlinear hybrid systems. Using this encoding, we apply a Satisfiability Modulo Theories (SMT) solver to find plans. We show that it is important to use novel planning- specific heuristics for variable and value selection for SMT solving, which is inspired by recent advances in planning as SAT. We show the promising performance of the resulting solver on challenging nonlinear problems.

Predictive state representations (PSRs) are models of dynamical systems that represent state as a vector of predictions about future observable events (tests) conditioned on past observed events (histories). If a practitioner selects finite sets of tests and histories that are known to be sufficient to completely capture the system, an exact PSR can be learned in polynomial time using spectral methods. However, most real-world systems are complex, and in practice computational constraints limit us to small sets of tests and histories which are therefore never truly sufficient. How, then, should we choose these sets? Existing theory offers little guidance here, and yet we show that the choice is highly consequential -- tests and histories selected at random or by a naive rule significantly underperform the best sets. In this paper we approach the problem both theoretically and empirically. While any fixed system can be represented by an infinite number of equivalent but distinct PSRs, we show that in the computationally unconstrained setting, where existing theory guarantees accurate predictions, the PSRs learned by spectral methods always satisfy a particular spectral bound. Adapting this idea, we propose a simple algorithmic technique to search for sets of tests and histories that approximately satisfy the bound while respecting computational limits. Empirically, our method significantly reduces prediction errors compared to standard spectral learning approaches.

However, it is well-known that applying machine learning and data mining to novel data sets is Finally, inspired by the field of constraint programming, challenging because each application imposes its own requirements (Guns et al. 2013) aim at developing declarative modeling and constraints that often require the development languages for specifying a wide range of mining problems. of new algorithms and systems. While there are software Such languages should support packages and tools such as Scikit for machine learning the high-level and natural modeling of pattern mining and Weka, Orange or Knime for data mining, adapting them tasks; that is, the models should closely correspond to to novel tasks is not easy, which explains why one often resorts the definitions of data mining problems found in the to implementing new algorithms and variations from literature; should support user-defined constraints and scratch.

This Blue Sky presentation focuses on a major shift toward a notion of “ambient intelligence” that transcends general applications targeted at the general population.  The focus is on highly personalized agents that accommodate individual differences and changes over time.  This notion of Extended Ambient Intelligence (EAI) concerns adaptation to a person’s preferences and experiences, as well as changing capabilities, most notably in an environment where conversational engagement is central.  An important step in moving this research forward is the accommodation of different degrees of cognitive capability (including speech processing) that may vary over time for a given user—whether through improvement or through deterioration. We suggest that the application of divergence detection to speech patterns may enable adaptation to a speaker’s increasing or decreasing level of speech impairment over time. Taking an adaptive approach toward technology development in this arena may be a first step toward empowering those with special needs so that they may live with a high quality of life.  It also represents an important step toward a notion of ambient intelligence that is personalized beyond what can be achieved by mass-produced, one-size-fits-all software currently in use on mobile devices.

Boolean satisfiability (SAT) solvers have been successfully applied to a wide variety of difficult combinatorial problems. Many further problems can be solved by SAT Modulo Theory (SMT) solvers, which extend SAT solvers to handle additional types of constraints. However, building efficient SMT solvers is often very difficult. In this paper, we define the concept of a Boolean monotonic theory and show how to easily build efficient SMT solvers, including effective theory propagation and clause learning, for such theories. We present examples showing useful constraints that are monotonic, including many graph properties (e.g., shortest paths), and geometric properties (e.g., convex hulls). These constraints arise in problems that are otherwise difficult for SAT solvers to handle, such as procedural content generation. We have implemented several monotonic theory solvers using the techniques we present in this paper and applied these to content generation problems, demonstrating major speed-ups over SAT, SMT, and Answer Set Programming solvers, easily solving instances that were previously out of reach.

This paper presents a large-scale sparse coding algorithm to deal with the challenging problem of noise-robust semi-supervised learning over very large data with only few noisy initial labels. By giving an L1-norm formulation of Laplacian regularization directly based upon the manifold structure of the data, we transform noise-robust semi-supervised learning into a generalized sparse coding problem so that noise reduction can be imposed upon the noisy initial labels. Furthermore, to keep the scalability of noise-robust semi-supervised learning over very large data, we make use of both nonlinear approximation and dimension reduction techniques to solve this generalized sparse coding problem in linear time and space complexity. Finally, we evaluate the proposed algorithm in the challenging task of large-scale semi-supervised image classification with only few noisy initial labels. The experimental results on several benchmark image datasets show the promising performance of the proposed algorithm.

Maximum entropy inverse optimal control (MaxEnt IOC) is an effective means of discovering the underlying cost function of demonstrated human activity and can be used to predict human behavior over low-dimensional state spaces (i.e., forecasting of 2D trajectories). To enable inference in very large state spaces, we introduce an approximate MaxEnt IOC procedure to address the fundamental computational bottleneck stemming from calculating the partition function via dynamic programming. Approximate MaxEnt IOC is based on two components: approximate dynamic programming and Monte Carlo sampling. We analyze this approximation approach and provide a finite-sample error upper bound on its excess loss. We validate the proposed method in the context of analyzing dual-agent interactions from video, where we use approximate MaxEnt IOC to simulate mental images of a single agents body pose sequence (a high-dimensional image space). We experiment with sequences image data taken from RGB and RGBD data and show that it is possible to learn cost functions that lead to accurate predictions in high-dimensional problems that were previously intractable.

In the Hierarchical Temporal Memory (HTM) paradigm the effect of overlap between inputs on the activation of columns in the spatial pooler is studied. Numerical results suggest that similar inputs are represented by similar sets of columns and dissimilar inputs are represented by dissimilar sets of columns. It is shown that the spatial pooler produces these results under certain conditions for the connectivity and proximal thresholds at initialization. Qualitative arguments about the learning dynamics of the spatial pooler are then discussed.