Publications

Publications
Publications
We strongly believe in open source and giving to our community. We work directly with researchers in academia and seek out new perspectives with our intern and fellowship programs. We generalize our solutions and release them to the world as open source projects. We host discussions and publish our results.

Publications

In IEEE VAST 2008 Symposium Challenge, 2008

Visual Analytics of Cell Phone Data using MobiVis and OntoVis

Carlos D.Correa, Tarik Crnovrsanin, Christopher Muelder, Zeqian Shen, Ryan Armstrong, James Shearer, Kwan-Liu Ma

MobiVis is a visual analytics tools to aid in the process of processing and understanding complex relational data, such as social networks. At the core of these tools is the ability to filter complex networks structurally and semantically, which helps us discover clusters and patterns in the organization of social networks.

Semantic filtering is obtained via an ontology graph, based on another visual analytics tool, called OntoVis. In this summary, we describe how these tools where used to analyze one of the mini-challenges of the 2008 VAST challenge.

Keywords
KDD 2008

Bypass rates: reducing query abandonment using negative inferences

Atish Das Sarma, Sreenivas Gollapudi, Samuel Ieong, Atish Das Sarma, Sreenivas Gollapudi, Samuel Ieong

We introduce a new approach to analyzing click logs by examining both the documents that are clicked and those that are bypassed-documents returned higher in the ordering of the search results but skipped by the user.

This approach complements the popular click-through rate analysis, and helps to draw negative inferences in the click logs. We formulate a natural objective that finds sets of results that are unlikely to be collectively bypassed by a typical user.

This is closely related to the problem of reducing query abandonment. We analyze a greedy approach to optimizing this objective, and establish theoretical guarantees of its performance.

We evaluate our approach on a large set of queries, and demonstrate that it compares favorably to the maximal marginal relevance approach on a number of metrics including mean average precision and mean reciprocal rank.

Keywords
Memory Studies, 09/2008, Volume 1, Issue 3, p.295-310, 2008

Technologies of Memory: Key Issues and Critical Perspectives

Nancy VanHouse, Elizabeth Churchill

Past, present and emerging technologies of memory are important concerns for memory studies. What is remembered individually and collectively depends in part on technologies of memory and socio-technical practices, which are changing radically. We identify specific concerns about developments in digital memory capture, storage and retrieval. Decisions are being made now that may have far-reaching consequences.

Systems are being designed based on models and metaphors in which human memory works much like the computer.

We bring to this discussion a critical perspective from science and technology studies (STS) and a grounding in human—computer interaction (HCI) and computer-supported cooperative work (CSCW).

We argue that, while these developments are significant for memory studies research, even more important is the need for memory studies to remind and inspire designers of what is possible and useful, and help expand the understanding of human memory on which these systems are based.

Keywords
In Proceedings of Eurographics/IEEE VGTC Syposium on Visualization, May 2007, pp. 83-90

Path Visualization for Adjacency Matrices

Zeqian Shen, Kwan-Liu Ma, Zeqian Shen, Kwan-Liu Ma

For displaying a dense graph, an adjacency matrix is superior than a node-link diagram because it is more compact and free of visual clutter. A node-link diagram, however, is far better for the task of path finding because a path can be easily traced by following the corresponding links, provided that the links are not heavily crossed or tangled.

We augment adjacency matrices with path visualization and associated interaction techniques to facilitate path finding.

Our design is visually pleasing, and also effectively displays multiple paths based on the design commonly found in metro maps. We illustrate and assess the key aspects of our design with the results obtained from two case studies and an informal user study.

Keywords
Categories
The Review of Agricultural Economics, 29(3):446-493. (2007)

The Productivity Argument for Investing in Young Children

James J.Heckman

This paper presents a productivity argument for investing in disadvantaged young children. For such investment, there is no equity-efficiency tradeoff.

Keywords
Categories
Accepted for presentation in the IEEE/RSJ International Conference on Intelligent Robots and Systems, 2007

A Supervised Learning Approach for Collision Detection in Legged Locomotion

Finale Doshi, Emma Brunskill, Alec Shkolnik, Thomas Kollar, Khashayar Rohanimanesh, Russ Tedrake, Nicholas Roy

We propose a fast approach for detecting collision-free swing-foot trajectories for legged locomotion over extreme terrains. Instead of simulating the swing trajectories and checking for collisions along them, our approach uses machine learning techniques to predict whether a swing trajectory is collision-free.

Using a set of local terrain features, we apply supervised learning to train a classifier to predict collisions. Both in simulation and on a real quadruped platform, our results show that our classifiers can improve the accuracy of collision detection compared to a real-time geometric approach without significantly increasing the computation time.

Keywords
Technical report, Massachusetts Institute of Technology Computer Science and Artificial Intelligence Laboratory, 2007

Towards feature selection in actor-critic algorithms

Khashayar Rohanimanesh, Russ Tedrake, Nick Roy, Khashayar Rohanimanesh, Russ Tedrake, Nick Roy

Choosing features for the critic in actor-critic algorithms with function approximation is known to be a challenge. Too few critic features can lead to degeneracy of the actor gradient, and too many features may lead to slower convergence of the learner.

In this paper, we show that a well studied class of actor policies satisfy the known requirements for convergence when the actor features are selected carefully. We demonstrate that two popular representations for value methods - the barycentric interpolators and the graph Laplacian proto-value functions - can be used to represent the actor in order to satisfy these conditions.

A consequence of this work is a generalization of the proto-value function methods to the continuous action actor-critic domain. Finally, we analyze the performance of this approach using a simulation of a torque-limited inverted pendulum.

Keywords

Pages