Very sparse random projections

By Ping Li, Trevor J. Hastie, Kenneth Ward Church. 2006

726 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Computational Mechanics, Computer Graphics and Computer-Aided Design, Statistics and Probability, Algorithms & Complexity

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computational Mechanics

In Computer Graphics and Computer-Aided Design

In Statistics and Probability

In Algorithms & Complexity

Citations in the corpus

Measuring the Intrinsic Dimension of Objective Landscapes (2018, methodology): “In fact, a method exists for “very sparse” random projections (Li et al., 2006), which achieves a density of 1 √ D .”
Failing Loudly: An Empirical Study of Methods for Detecting Dataset Shift (2018, methodology): “The elements of R are generated using the following rule set [1, 27]:”
Compact Bilinear Pooling (2015, background): “A sparse random projection matrix would improve the memory consumption to around 40MB[21], but would still requires forming bilinear feature first.”
A Survey on Learning to Hash (2016, background): “computation of hash functions [71], [74], [54], [130].”
Hashing for Similarity Search: A Survey (2014, background): “This part of the study mainly focuses on developing variants of LSH functions that provide an (unbiased) estimator of certain distance or similarity, with smaller variance [68], [52], [73], [51], or smaller storage requirement of the hash codes [70], [71], or faster computation of hash functions [69”
Data-Driven Sparse Sensor Placement for Reconstruction: Demonstrating the Benefits of Exploiting Known Patterns (2017, background): “In addition, there is additional theory about how sparse the random projections may be for compressed sensing [111], [112].”
Deep Learning for Anomaly Detection (2020, methodology): “w space to calculate anomaly scores. Compared to the dimension reduction methods that are popular in anomaly detection, such as principal component analysis (PCA) [21, 141, 180] and random projection [81, 114, 124], deep learning techniques have been demonstrating substantially better capability in ”
Evaluation methods for unsupervised word embeddings (2015, methodology): “Second we use a recently released dataset for binary sentiment classification by Maas et al. (2011). The dataset contains 50K movie reviews with a balanced distribution of binary polarity labels.”
A Survey of Deep Learning Methods for Cyber Security (2019, methodology): “[95] combined feature selection and random projections [96,97] to reduce the dimensionality of their dataset to produce DNNs used for classifying malware.”
An Analysis of Random Projection for Changeable and Privacy-Preserving Biometric Verification (2010, background): “[29] showed that it is possible to achieve such embedding through much simpler random matrices for fast operation.”
Recent Advances of Large-Scale Linear Classification (2012, background): “Examples include random Fourier projection [77], random projections [78], [79], polynomial approximation [80], and hashing [81]– [84].”
Real-Time Compressive Tracking (2012, background): “[19] showed that for s = O(m) (x ∈ R), this matrix is asymptotically normal.”
A Multi-View Deep Learning Approach for Cross Domain User Modeling in Recommendation Systems (2015, background): “A lot of them are based on generating sparse A [13].”
A survey of randomized algorithms for training neural networks (2016, methodology): “) on different neural network classes by using the compressive sensing theory [13, 45, 14, 4].”
Sketch-Based Linear Value Function Approximation (2012, background): “random projection methods [1, 14] carry a per-feature cost that increases with the size of the reduced space.”
One Permutation Hashing (2012, methodology): “We are also inspired by the work on very sparse random projections [15] and very sparse stable random projections [12].”
RAPTT: An Exact Two-Sample Test in High Dimensions Using Random Projections (2014, methodology): “Our second choice of R is based on the idea of one permutation + one random projection , which is closely related to very sparse random projection (Li et al., 2006) and count-sketch (Charikar et al., 2004).”