Active object exploration is one of the hallmarks of human and animal intelligence. Research in psychology has shown that the use of multiple exploratory behaviors is crucial for learning about objects. Inspired by such research, recent work in robotics has demonstrated that by performing multiple exploratory behaviors a robot can dramatically improve its object recognition rate. But what is the cause of this improvement? To answer this question, this paper examines the conditions under which combining information from multiple behaviors and sensory modalities leads to better object recognition results. Two different problems are considered: interactive object recognition using auditory and proprioceptive feedback, and surface texture recognition using tactile and proprioceptive feedback. Analysis of the results shows that metrics designed to estimate classifier model diversity can explain the improvement in recognition accuracy. This finding establishes, for the first time, an important link between empirical studies of exploratory behaviors in robotics and theoretical results on boosting in machine learning.

Multi-task learning aims at combining information across tasks to boost prediction performance, especially when the number of training samples is small and the number of predictors is very large. In this paper, we first extend the Sparse Discriminate Analysis (SDA) of Clemmensen et al.. We call this Multi-task Sparse Discriminate Analysis (MtSDA). MtSDA formulates multi-label prediction as a quadratic optimization problem whereas SDA obtains single labels via a nearest class mean rule. Second, we propose a class of equicorrelation matrices to use in MtSDA which includes the identity matrix. MtSDA with both matrices are compared with singletask learning (SVM and LDA+SVM) and multi-task learning (HSML). The comparisons are made on real data sets in terms of AUC and F-measure. The data results show that MtSDA outperforms other methods substantially almost all the time and in some cases MtSDA with the equicorrelation matrix substantially outperforms MtSDA with identity matrix.

Knowledge transfer has been suggested as a useful approach for solving large Markov Decision Processes. The main idea is to compute a decision-making policy in one environment and use it in a different environment, provided the two are ”close enough”. In this paper, we use bisimulation-style metrics (Ferns et al., 2004) to guide knowledge transfer. We propose algorithms that decide what actions to transfer from the policy computed on a small MDP task to a large task, given the bisimulation distance between states in the two tasks. We demonstrate the inherent ”pessimism” of bisimulation metrics and present variants of this metric aimed to overcome this pessimism, leading to improved action transfer. We also show that using this approach for transferring temporally extended actions (Sutton et al., 1999) is more successful than using it exclusively with primitive actions. We present theoretical guarantees on the quality of the transferred policy, as well as promising empirical results.

Within the electric power literature the transmission expansion planning problem (TNEP) refers to the problem of how to upgrade an electric power network to meet future demands. As this problem is a complex, non-linear, and non-convex optimization problem, researchers have traditionally focused on approximate models of power flows. Existing approaches are often tightly coupled to the approximation choice. Until recently, these approximations have produced results that are straight-forward to adapt to the more complex (real) problem. However, the power grid is evolving towards a state where the adaptations are no longer easy (e.g. large amounts of limited control, renewable generation) that necessitates new optimization techniques. In this paper, we propose a local search variation of the powerful Limited Discrepancy Search (LDLS) that encapsulates the complexity of power flows in a black box that may be queried for information about the quality of a proposed expansion. This allows the development of a new optimization algorithm that is independent of the underlying power model.

We consider here the problem of building a never-ending language learner; that is, an intelligent computer agent that runs forever and that each day must (1) extract, or read, information from the web to populate a growing structured knowledge base, and (2) learn to perform this task better than on the previous day. In particular, we propose an approach and a set of design principles for such an agent, describe a partial implementation of such a system that has already learned to extract a knowledge base containing over 242,000 beliefs with an estimated precision of 74% after running for 67 days, and discuss lessons learned from this preliminary attempt to build a never-ending learning agent.

We consider planning domains with both discrete and continuous changes. Continuous change occurs especially when agents share time-dependent critical resources. In these cases, besides discrete and continuous changes, their interactions should also be taken into consideration. However concurrency of durative actions with interacting continuous effects cannot be exploited by existing temporal planners. To overcome this problem, we propose an action lifting approach and we analyze path sharing problem to illustrate interaction of continuous linear effects in the planning domain.

This paper presents CAO, a system for affect analysis of emoticons. Emoticons are strings of symbols widely used in text-based online communication to convey emotions. It extracts emoticons from input and determines specific emotions they express. Firstly, by matching the extracted emoticons to a raw emoticon database, containing over ten thousand emoticon samples extracted from the Web and annotated automatically. The emoticons for which emotion types could not be determined using only this database, are automatically divided into semantic areas representing "mouths" or "eyes," based on the theory of kinesics. The areas are automatically annotated according to their co-occurrence in the database. The annotation is firstly based on the eye-mouth-eye triplet, and if no such triplet is found, all semantic areas are estimated separately. This provides the system coverage exceeding 3 million possibilities. The evaluation, performed on both training and test sets, confirmed the system's capability to sufficiently detect and extract any emoticon, analyze its semantic structure and estimate the potential emotion types expressed. The system achieved nearly ideal scores, outperforming existing emoticon analysis systems.

Policy search is a successful approach to reinforcement learning. However, policy improvements often result in the loss of information. Hence, it has been marred by premature convergence and implausible solutions. As first suggested in the context of covariant policy gradients, many of these problems may be addressed by constraining the information loss. In this paper, we continue this path of reasoning and suggest the Relative Entropy Policy Search (REPS) method. The resulting method differs significantly from previous policy gradient approaches and yields an exact update step. It can be shown to work well on typical reinforcement learning benchmark problems.

This paper presents a machine learning approach for accurate estimation of energy expenditure using a fusion of accelerometer and heart rate sensing. To address short comings in existing off-the-shelf solutions, we designed Jog Falls, an end to end system for weight management in collaboration with physicians in India. This system is meant to enable people to accurately monitor their energy expenditure and intake and make educated tradeoffs to reach their weight goals. In this paper we describe the sensing components of Jog Falls and focus on the energy expenditure estimation algorithm. We present results from controlled experiments in the lab, as well results from a 15 participant user study over a period of 63 days. We show how our algorithm mitigates many of the issues in existing solutions and yields more accurate results.

Modeling human behavior requires vast quantities of accurately labeled training data, but for ubiquitous people-aware applications such data is rarely attainable. Even researchers make mistakes when labeling data, and consistent, reliable labels from low-commitment users are rare. In particular, users may give identical labels to activities with characteristically different signatures (e.g., labeling eating at home or at a restaurant as "dinner") or may give different labels to the same context (e.g., "work" vs. "office"). In this scenario, labels are unreliable but nonetheless contain valuable information for classification. To facilitate learning in such unconstrained labeling scenarios, we propose Community-Guided Learning (CGL), a framework that allows existing classifiers to learn robustly from unreliably-labeled user-submitted data. CGL exploits the underlying structure in the data and the unconstrained labels to intelligently group crowd-sourced data. We demonstrate how to use similarity measures to determine when and how to split and merge contributions from different labeled categories and present experimental results that demonstrate the effectiveness of our framework.