Skip to content
conftrace_
2018 ICML ICML 2018

Streaming Principal Component Analysis in Noisy Setting

Abstract

We study streaming algorithms for principal component analysis (PCA) in noisy settings. We present computationally efficient algorithms with sub-linear regret bounds for PCA in the presence of noise, missing data, and gross outliers.

🐣 Hot Topic Early Bird - sublinear regret

Authors

Teodor Vanislavov Marinov, Poorya Mianjy, Raman Arora

Topics

Machine Learning > Core Methods > Representation Learning Machine Learning > Core Methods > Dimensionality Reduction Machine Learning > Learning Types > Online Learning Machine Learning > Optimization & Theory > Online Algorithms

Keywords

principal component analysis outlier detection streaming principal component analysis streaming algorithm sublinear regret missing data noisy setting
Download PDF

More from ICML 2018

All ICML 2018 papers →
Learning Adversarially Fair and Transferable Representations
An Efficient, Generalized Bellman Update For Cooperative Inverse Reinforcement Learning
Iterative Amortized Inference
Fast Approximate Spectral Clustering for Dynamic Networks
Bayesian Model Selection for Change Point Detection and Clustering
Optimization, fast and slow: optimally switching between local and Bayesian optimization