Deflation Methods for Sparse PCA
- Lester W. Mackey
- Computer ScienceNIPS
- 8 December 2008
This work develops several deflation alternatives better suited to the cardinality-constrained context and reformulates the sparse PCA optimization problem to explicitly reflect the maximum additional variance objective on each round, resulting in a generalized deflation procedure that typically outperforms more standard techniques on real-world datasets.
Minimax Estimation of Conditional Moment Models
- Nishanth Dikkala, Greg Lewis, Lester W. Mackey, Vasilis Syrgkanis
- Mathematics, Computer ScienceNeural Information Processing Systems
- 12 June 2020
This work develops an approach for estimating models described via conditional moment restrictions, and introduces a min-max criterion function, under which the estimation problem can be thought of as solving a zero-sum game between a modeler who is optimizing over the hypothesis space of the target model and an adversary who identifies violating moments over a test function space.
Measuring Sample Quality with Kernels
- Jackson Gorham, Lester W. Mackey
- MathematicsInternational Conference on Machine Learning
- 6 March 2017
A theory of weak convergence for K SDs based on Stein's method is developed, it is demonstrated that commonly used KSDs fail to detect non-convergence even for Gaussian targets, and it is shown that kernels with slowly decaying tails provably determine convergence for a large class of target distributions.
Divide-and-Conquer Matrix Factorization
- Lester W. Mackey, Ameet S. Talwalkar, Michael I. Jordan
- Computer ScienceNIPS
- 5 July 2011
The experiments with collaborative filtering, video background modeling, and simulated data demonstrate the near-linear to super-linear speed-ups attainable with DFC, and the analysis shows that DFC enjoys high-probability recovery guarantees comparable to those of its base algorithm.
Measuring Sample Quality with Stein's Method
- Jackson Gorham, Lester W. Mackey
- Computer ScienceNIPS
- 9 June 2015
This work introduces a new computable quality measure based on Stein's method that quantifies the maximum discrepancy between sample and target expectations over a large class of test functions and uses this tool to compare exact, biased, and deterministic sample sequences.
Matrix concentration inequalities via the method of exchangeable pairs
- Lester W. Mackey, Michael I. Jordan, Richard Y. Chen, Brendan Farrell, J. Tropp
- Mathematics
- 28 January 2012
This paper derives exponential concentration inequalities and polynomial moment inequalities for the spectral norm of a random matrix. The analysis requires a matrix extension of the scalar…
Feature-Weighted Linear Stacking
- J. Sill, G. Takács, Lester W. Mackey, David Lin
- Computer ScienceArXiv
- 3 November 2009
A linear technique, Feature-Weighted Linear Stacking (FWLS), that incorporates meta-features for improved accuracy while retaining the well-known virtues of linear regression regarding speed, stability, and interpretability is presented.
Corrupted Sensing: Novel Guarantees for Separating Structured Signals
- Rina Foygel, Lester W. Mackey
- Computer ScienceIEEE Transactions on Information Theory
- 11 May 2013
This work analyzes both penalized programs that tradeoff between signal and corruption complexity, and constrained programs that bound the complexity of signal or corruption when prior information is available, and provides new interpretable bounds for the Gaussian complexity of sparse vectors, block-sparse vectors, and low-rank matrices.
Crowdsourced analysis of clinical trial data to predict amyotrophic lateral sclerosis progression
- R. Küffner, N. Zach, Melanie L Leitner
- PsychologyNature Biotechnology
- 2015
The DREAM-Phil Bowen ALS Prediction Prize4Life challenge identified several potential nonstandard predictors of disease progression including uric acid, creatinine and surprisingly, blood pressure, shedding light on ALS pathobiology.
Joint Link Prediction and Attribute Inference Using a Social-Attribute Network
- N. Gong, Ameet S. Talwalkar, D. Song
- Computer ScienceTIST
- 1 April 2014
The novel observation that attribute inference can help inform link prediction, that is, link-prediction accuracy is further improved by first inferring missing attributes, is made.
...
...