I present a robotics localization challenge based on the inexpensive Neato XV robotic vacuum cleaner platform. The challenge teaches skills such as computational modeling, probabilistic inference, efficiency vs. accuracy tradeoffs, debugging, parameter tuning, and benchmarking of algorithmic performance. Rather than allowing students to pursue any localization algorithm of their choosing, here, I propose a challenge structured around the particle filter family of algorithms. This additional scaffolding allows students at all levels to successfully implement one approach to the challenge, while providing enough flexibility and richness to enable students to pursue their own creative ideas. Additionally, I provide infrastructure for automatic evaluation of systems through the collection of ground truth robot location data via ceiling-mounted location tags that are automatically scanned using an upward facing camera attached to the robot. The robot and supporting hardware can be purchased for under $400 dollars, and the challenge can even be run without any robots at all using a set of recorded sensor traces.

One of the main goals of natural language processing (NLP) is synthetic understanding of natural language documents, especially reading comprehension (RC). An obstacle to the further development of RC systems is the absence of a synthetic methodology to analyze their performance. It is difficult to examine the performance of systems based solely on their results for tasks because the process of natural language understanding is complex. In order to tackle this problem, we propose in this paper a methodology inspired by unit testing in software engineering that enables the examination of RC systems from multiple aspects. Our methodology consists of three steps. First, we define a set of prerequisite skills for RC based on existing NLP tasks. We assume that RC capability can be divided into these skills. Second, we manually annotate a dataset for an RC task with information regarding the skills needed to answer each question. Finally, we analyze the performance of RC systems for each skill based on the annotation. The last two steps highlight two aspects: the characteristics of the dataset, and the weaknesses in and differences among RC systems. We tested the effectiveness of our methodology by annotating the Machine Comprehension Test (MCTest) dataset and analyzing four existing systems (including a neural system) on it. The results of the annotations showed that answering questions requires a combination of skills, and clarified the kinds of capabilities that systems need to understand natural language. We conclude that the set of prerequisite skills we define are promising for the decomposition and analysis of RC.

This paper studies a criteria-based mechanism for nurturing and enhancing agents' group-benefiting individual efforts whenever the agents are self-interested. The idea is that only those agents that meet the criteria get to benefit from the group effort, giving an incentive to contribute even when it is otherwise individually irrational. Specifically, the paper provides a comprehensive equilibrium analysis of a threshold-based criteria mechanism for the common cooperative information gathering application, where the criteria is set such that only those whose contribution to the group is above some pre-specified threshold can benefit from the contributions of others. The analysis results in a closed form solution for the strategies to be used in equilibrium and facilitates the numerical investigation of different model properties as well as a comparison to the dual mechanism according to only an agent whose contribution is below the specified threshold gets to benefit from the contributions of others. One important contribution enabled through the analysis provided is in showing that, counter-intuitively, for some settings the use of the above-threshold criteria is outperformed by the use of the below-threshold criteria as far as collective and individual performance is concerned.

Matrix factorization is a core technique in many machine learning problems, yet also presents a nonconvex and often difficult-to-optimize problem. In this paper we present an approach based upon polynomial optimization techniques that both improves the convergence time of matrix factorization algorithms and helps them escape from local optima. Our method is based on the realization that given a joint search direction in a matrix factorization task, we can solve the ``subspace search'' problem (the task of jointly finding the steps to take in each direction) by solving a bivariate quartic polynomial optimization problem. We derive two methods for solving this problem based upon sum of squares moment relaxations and the Durand-Kerner method, then apply these techniques on matrix factorization to derive a direct coordinate descent approach and a method for speeding up existing approaches. On three benchmark datasets we show the method substantially improves convergence speed over state-of-the-art approaches, while also attaining lower objective value.

We study the problem of structured prediction under test-time budget constraints. We propose a novel approach based on selectively acquiring computationally costly features during test-time in order to reduce the computational cost of pre- diction with minimal performance degradation. We formulate a novel empirical risk minimization (ERM) for policy learning. We show that policy learning can be reduced to a series of structured learning problems, resulting in efficient training using existing structured learning algorithms. This framework provides theoretical justification for several existing heuristic approaches found in literature. We evaluate our proposed adaptive system on two structured prediction tasks, optical character recognition and dependency parsing and show significant reduction in the feature costs without degrading accuracy.

In this paper we present an on-manifold sequence-to-sequence learning approach to motion estimation using visual and inertial sensors. It is to the best of our knowledge the first end-to-end trainable method for visual-inertial odometry which performs fusion of the data at an intermediate feature-representation level. Our method has numerous advantages over traditional approaches. Specifically, it eliminates the need for tedious manual synchronization of the camera and IMU as well as eliminating the need for manual calibration between the IMU and camera. A further advantage is that our model naturally and elegantly incorporates domain specific information which significantly mitigates drift. We show that our approach is competitive with state-of-the-art traditional methods when accurate calibration data is available and can be trained to outperform them in the presence of calibration and synchronization errors.

Learning fast and efficiently using minimal data has been consistently a challenge in machine learning. In my thesis, I explore this problem for knowledge transfer for multi-agent multi-task learning in a life-long learning paradigm. My goal is to demonstrate that by sharing knowledge between agents and similar tasks, efficient algorithms can be designed that can increase the speed of learning as well as improve performance. Moreover, this would allow for handling hard tasks through collective learning of multiple agents that share knowledge. As an initial step, I study the problem of incorporating task descriptors into lifelong learning of related tasks to perform zero-shot knowledge transfer. Zero-shot learning is highly desirable because it leads to considerable speedup in handling similar sequential tasks. Then I focus on a multi-agent learning setting, where related tasks are learned collectively and/or address privacy concerns.

Systems deployed in unstructured environments must be able to adapt to novel situations. This requires the ability to perform in domains that may be vastly different from training domains. My dissertation focuses on the representations used in lifelong learning and how these representations enable predictions and knowledge sharing over time, allowing an agent to continuously learn and adapt in changing environments. Specifically, my contributions will enable lifelong learning systems to efficiently accumulate data, use prior knowledge to predict models for novel tasks, and alter existing models to account for changes in the environment.

Reservoir Computing is a bio-inspired computing paradigm for processing time-dependent signals (Jaeger and Haas 2004; Maass, Natschläger, and Markram 2002). The performance of its hardware implementation (see e.g. (Soriano et al. 2015) for a review) is comparable to state-of-the-art digital algorithms on a series of benchmark tasks.The major bottleneck of these implementation is the readout layer, based on slow offline post-processing. Several analogue solutions have been proposed (Smerieri et al. 2012; Duport et al. 2016; Vinckier et al. 2016), but all suffered from noticeable decrease in performance due to added complexity of the setup. Here we propose the online learning approach to solve these issues. We present an experimental reservoir computer with a simple analogue readout layer, based on previous works, and show numerically that online learning allows to disregard the added complexity of an analogue layer and obtain the same level of performance as with a digital layer. This work thus demonstrates that online training allows building high-performance fully-analogue reservoir computers, and represents an important step towards experimental validation of the proposed solution.

The Model AI Assignments session seeks to gather and disseminate the best assignment designs of the Artificial Intelligence (AI) Education community. Recognizing that assignments form the core of student learning experience, we here present abstracts of six AI assignments from the 2017 session that are easily adoptable, playfully engaging, and flexible for a variety of instructor needs.