Share This Author
Random Features for Large-Scale Kernel Machines
Two sets of random features are explored, provided convergence bounds on their ability to approximate various radial basis kernels, and it is shown that in large-scale classification and regression tasks linear machine learning algorithms applied to these features outperform state-of-the-art large- scale kernel machines.
Similarity-based Classification: Concepts and Algorithms
- Yihua Chen, E. K. Garcia, M. Gupta, A. Rahimi, L. Cazzanti
- Computer ScienceJ. Mach. Learn. Res.
- 1 December 2009
The generalizability of using similarities as features is analyzed, design goals and methods for weighting nearest-neighbors for similarity-based learning are proposed, and different methods for consistently converting similarities into kernels are compared.
Uniform approximation of functions with random bases
- A. Rahimi, B. Recht
- Computer Science46th Annual Allerton Conference on Communication…
- 1 September 2008
Using techniques from probability on Banach Spaces, a specific architecture of random nonlinearities is analyzed, Linfin and L2 error bounds for approximating functions in Reproducing Kernel Hilbert Spaces are provided and scenarios when these expansions are dense in the continuous functions are discussed.
Weighted Sums of Random Kitchen Sinks: Replacing minimization with randomization in learning
Shallow random networks are analyzed, which are architectures that compute a weighted sum of their inputs after passing them through a bank of arbitrary randomized nonlinearities, and bound their test error in terms of the size of the dataset and the number of random non linearities.
Simultaneous calibration and tracking with a network of non-overlapping sensors
A method for simultaneously recovering the trajectory of a target and the external calibration parameters of non-overlapping cameras in a multi-camera system with a network of indoor wireless cameras is described.
3D pose tracking with linear depth and brightness constraints
- M. Harville, A. Rahimi, Trevor Darrell, G. Gordon, J. Woodfill
- MathematicsProceedings of the Seventh IEEE International…
- 20 September 1999
This paper explores the direct motion estimation problem assuming that video-rate depth information is available, from either stereo cameras or other sensors, and derives linear brightness and depth change constraint equations that govern the velocity field in 3D for both perspective and orthographic camera projection models.
Simultaneous localization, calibration, and tracking in an ad hoc sensor network
- Christopher Taylor, A. Rahimi, J. Bachrach, H. Shrobe, Anthony Grue
- Computer Science5th International Conference on Information…
- 19 April 2006
The proposed solution, LaSLAT, is a Bayesian filter that provides on-line probabilistic estimates of sensor locations and target tracks that does not require globally accessible beacon signals or accurate ranging between the nodes.
Adaptive view-based appearance models
- Louis-Philippe Morency, A. Rahimi, Trevor Darrell
- Computer ScienceIEEE Computer Society Conference on Computer…
- 18 June 2003
This work presents a method for online rigid object tracking using an adaptive view-based appearance model that has bounded drift and can track objects undergoing large motion for long periods of time when the object's pose trajectory crosses itself.
Clustering with Normalized Cuts is Clustering with a Hyperplane
It turns out that the Normalized Cuts algorithm, originally presented as a graph-theoretic algorithm, can be interpreted as an algorithm that assigns uniform weight to all points, eliminating the sensitivity to outliers.
Semi-Supervised Learning with Max-Margin Graph Cuts
This paper proposes a novel algorithm that learns graph cuts that maximize the margin with respect to the labels induced by the harmonic function solution that outperforms manifold regularization of support vector machines.