Feature Selection for Automatic Classification of Non-Gaussian Data

Foroutan, I.; Sklansky, Jack

doi:10.1109/tsmc.1987.4309029

Cited by 68 publications

(23 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The FS-K Means_BIC uses SFS [14] to search for feature subsets. The criterion used in this algorithm is the traceðS À1 W S B Þ normalized using a cross projection scheme [12].…”

Section: Resultsmentioning

confidence: 99%

“…Sequential forward selection (SFS) and sequential backward selection (SBS) [14] are two classical heuristic feature selection algorithms developed for supervised learning. SFS starts with an empty set of features, and at each iteration, the algorithm tentatively adds each available feature and selects the feature that results in the highest estimated performance.…”

Section: Local Searchesmentioning

confidence: 99%

See 1 more Smart Citation

A Niching Memetic Algorithm for Simultaneous Clustering and Feature Selection

Sheng

Liu²,

Fairhurst

2008

IEEE Trans. Knowl. Data Eng.

View full text Add to dashboard Cite

Abstract-Clustering is inherently a difficult task and is made even more difficult when the selection of relevant features is also an issue. In this paper, we propose an approach for simultaneous clustering and feature selection using a niching memetic algorithm. Our approach (which we call NMA_CFS) makes feature selection an integral part of the global clustering search procedure and attempts to overcome the problem of identifying less promising locally optimal solutions in both clustering and feature selection, without making any a priori assumption about the number of clusters. Within the NMA_CFS procedure, a variable composite representation is devised to encode both feature selection and cluster centers with different numbers of clusters. Furthermore, local search operations are introduced to refine feature selection and cluster centers encoded in the chromosomes. Finally, a niching method is integrated to preserve the population diversity and prevent premature convergence. In an experimental evaluation, we demonstrate the effectiveness of the proposed approach by using both synthetic and real data.

show abstract

“…The FS-K Means_BIC uses SFS [14] to search for feature subsets. The criterion used in this algorithm is the traceðS À1 W S B Þ normalized using a cross projection scheme [12].…”

Section: Resultsmentioning

confidence: 99%

Section: Local Searchesmentioning

confidence: 99%

A Niching Memetic Algorithm for Simultaneous Clustering and Feature Selection

Sheng

Liu²,

Fairhurst

2008

IEEE Trans. Knowl. Data Eng.

View full text Add to dashboard Cite

show abstract

“…Moreover, even if 99.9% computational cost saving is done by BB, the computational complexity remains exponential. Thus, BB or its improved version*relaxed BB (by introducing the concept of approximate monotonicity) [3]*are still unavailable for a large n.…”

Section: T S-f Error(s)(tmentioning

confidence: 99%

Feature selection using tabu search method

Zhang

Sun

2002

Pattern Recognition

226

107

View full text Add to dashboard Cite

Selecting an optimal subset from original large feature set in the design of pattern classi"er is an important and di$cult problem. In this paper, we use tabu search to solve this feature selection problem and compare it with classic algorithms, such as sequential methods, branch and bound method, etc., and most other suboptimal methods proposed recently, such as genetic algorithm and sequential forward (backward) #oating search methods. Based on the results of experiments, tabu search is shown to be a promising tool for feature selection in respect of the quality of obtained feature subset and computation e$ciency. The e!ects of parameters in tabu search are also analyzed by experiments.

show abstract

“…If u ≥ 1 2 then choose a model in NB IG(+1) with the probability |NB IG(+1) | −1 , otherwise choose a model in NB IG(−1) with the probability |NB IG(−1) | −1 . The normalization factor that ensures detailed balance, becomes…”

Section: Naive Bayes Classifiersmentioning

confidence: 99%

“…Important key references include [1][2][3][4]. Whether one wants to learn a statistical classifier from data, a graphical model or perform clustering in high-dimensional space, confining the number of feature variables included in the model by either feature selection or feature transformation, is often necessary.…”

Section: Introductionmentioning

confidence: 99%

Feature Selection by Markov Chain Monte Carlo Sampling – A Bayesian Approach

Egmont-Petersen

2004

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Abstract. We redefine the problem of feature selection as one of model selection and propose to use a Markov Chain Monte Carlo method to sample models. The applicability of our method is related to Bayesian network classifiers. Simulation experiments indicate that our novel proposal distribution results in an ignorant proposal prior. Finally, it is shown how the sampling can be controlled by a regularization prior.

show abstract

Feature Selection for Automatic Classification of Non-Gaussian Data

Cited by 68 publications

References 23 publications

A Niching Memetic Algorithm for Simultaneous Clustering and Feature Selection

A Niching Memetic Algorithm for Simultaneous Clustering and Feature Selection

Feature selection using tabu search method

Feature Selection by Markov Chain Monte Carlo Sampling – A Bayesian Approach

Contact Info

Product

Resources

About