A filter-dominating hybrid sequential forward floating search method for feature subset selection in high-dimensional space

Gan, John Q.; Hasan, Bashar; Tsui, Chun Sing Louis

doi:10.1007/s13042-012-0139-z

Cited by 30 publications

(16 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For wrapper approach, the features were selected using the sequential forward floating search (SFFS) method with LDA classifier as the wrapper. [29] …”

Section: Methodsmentioning

confidence: 99%

Re-ranking Google search returned web documents using document classification scores

Plansangket

Gan

2016

AIR

View full text Add to dashboard Cite

Web document ranking is a very challenging issue for search engines because about 80% of the search engine users are usually interested in the top three returned search results only. This paper proposes an effective method for re-ranking Google search returned web documents/pages based on document classification. This method downgrades some web documents/pages that have lower classification scores or been classified into categories irrelevant to the query. The experimental results show that the re-ranking of Google search returned web documents using document classification scores can significantly improve the ranking performance in terms of the integrated evaluation result using three criteria: MAP, nDCG, and P@20. It is evident that the proposed re-ranking method can meet the user's information need better.

show abstract

“…For wrapper approach, the features were selected using the sequential forward floating search (SFFS) method with LDA classifier as the wrapper. [29] …”

Section: Methodsmentioning

confidence: 99%

Re-ranking Google search returned web documents using document classification scores

Plansangket

Gan

2016

AIR

View full text Add to dashboard Cite

show abstract

“…A common form of this is a two-stage approach: a filter method is first applied to remove the most redundant features, before a wrapper is applied to the remaining features. A variation of this is seen in Gan [12], where Sequential Forward Floating Search (SFFS) was combined with MRMR by using the mutual information approach to select a set of candidate features for addition and removal at each phase. This reduced the computational training cost of utilising the classifier across all the candidate features.…”

Section: Hybrid Approachesmentioning

confidence: 99%

Mutual Information Iterated Local Search: A Wrapper-Filter Hybrid for Feature Selection in Brain Computer Interfaces

Adair

Brownlee

Ochoa

2018

Applications of Evolutionary Computation

View full text Add to dashboard Cite

Brain Computer Interfaces provide a very challenging classification task due to small numbers of instances, large numbers of features, non-stationary problems, and low signal-to-noise ratios. Feature selection (FS) is a promising solution to help mitigate these effects. Wrapper FS methods are typically found to outperform filter FS methods, but reliance on cross-validation accuracies can be misleading due to overfitting. This paper proposes a filter-wrapper hybrid based on Iterated Local Search and Mutual Information, and shows that it can provide more reliable solutions, where the solutions are more able to generalise to unseen data. This study further contributes comparisons over multiple datasets, something that has been uncommon in the literature.

show abstract

“…Below we assess recent developments in solutions for high-dimensionality problems in areas such as clustering [33,34], regression [35,36,37] and classification [38,39].…”

Section: Recent Contributionsmentioning

confidence: 99%

Recent advances and emerging challenges of feature selection in the context of big data

Bolón-Canedo

Sánchez-Maroño

Alonso‐Betanzos

2015

Knowledge-Based Systems

253

113

View full text Add to dashboard Cite

A filter-dominating hybrid sequential forward floating search method for feature subset selection in high-dimensional space

Cited by 30 publications

References 15 publications

Re-ranking Google search returned web documents using document classification scores

Re-ranking Google search returned web documents using document classification scores

Mutual Information Iterated Local Search: A Wrapper-Filter Hybrid for Feature Selection in Brain Computer Interfaces

Recent advances and emerging challenges of feature selection in the context of big data

Contact Info

Product

Resources

About