Feature Selection for Unsupervised Learning

By Jennifer G. Dy, Carla E. Brodley. Journal of machine learning research, 2004

1,094 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/1e4558354f9509ab9992001340f88be9f298debe

How researchers see this paper

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

1. Yi Yang
2. Dacheng Tao
3. Feiping Nie
4. Luc Van Gool
5. Shuicheng Yan
6. Kristen Grauman
7. Deng Cai
8. Trevor Darrell

Citations in the corpus

• The Chosen Few: On Identifying Valuable Patterns (2007, background): “An example that seems very close to our work is [5] in that the wrapper uses unsupervised learning instead of an induction algorithm.”
• Spectral feature selection for supervised and unsupervised learning (2007, methodology): “And we show that existing power­ful algorithms such as ReliefF (supervised) and Laplacian Score (unsupervised) are spe­cial cases of the proposed framework.”
• Artificial Intelligence for Mental Health and Mental Illnesses: An Overview (2019, methodology): “UML uses clustering techniques (e.g., k-means, hierarchical, principal component analysis) to sort and separate data into groups or patterns or identify the most salient features of a dataset [44].”
• Feature Selection for Classification: A Review (2014, background): “According to whether the training set is labelled or not, feature selection algorithms can be categorized into supervised [68, 61], unsupervised [13, 51] and semi-supervised feature selection [77, 71].”
• Discriminative Semi-Supervised Feature Selection Via Manifold Regularization (2009, background): “feature subset that best keeps the intrinsic clusters from data according to the specified clustering criterion [17], [53].”
• One in a million: picking the right patterns (2009, background): “An example that seems very close to our work is [7] in that the wrapper uses unsupervised learning instead of an induction algorithm.”
• Mind the Gap: A Generative Approach to Interpretable Feature Selection and Extraction (2015, background): “[16]) iterate between selecting dimensions and clustering to maximize a clustering objective.”
• Foreground Focus: Unsupervised Learning from Partially Matching Images (2009, background): “The problem of unsupervised feature selection has received limited attention in the machine learning community (see Dy and Brodley 2004 and references therein), but existing methods presume a vector input space, many assume the data to be generated by certain parametric distributions, and/or are s”
• Supervised, Unsupervised, and Semi-Supervised Feature Selection: A Review on Gene Selection (2016, background): “4a and 4b show the examples of redundant and irrelevant features [160].”
• Unsupervised feature selection for principal components analysis (2008, background): “[16] presents an expectationmaximization based scheme for unsupervised feature selection that comes with some provable guarantees under assumptions.”
• Unsupervised feature selection by regularized self-representation (2015, methodology): “Feature selection methods can be categorized into unsupervised and supervised ones [4,5,13].”
• Clustering-Guided Sparse Structural Learning for Unsupervised Feature Selection (2014, methodology): “Wrapper methods [4], [11], [26] use a predictive model to score feature subsets.”
• Feature Selection for Clustering: A Review (2018, methodology): “The method proposed in [18] is an example of a wrapper that involves maximum likelihood criteria and feature selection and mixture of Gaussians as clustering method.”
• Selection of Variables for Cluster Analysis and Classification Rules (2006, unknown): “In addition, they concluded that the mixture modeling techniques, such as feature saliency (Law, Figueiredo, and Jain 2004), RD, and scatter separability (Dy and Brodley 2004), performed at a very disappointing level.”