High dimensional data classification and feature selection using support vector machines

Ghaddar, Bissan; Naoum-Sawaya, Joe

doi:10.1016/j.ejor.2017.08.040

Cited by 218 publications

(115 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…For each dataset, a subset of positive and negative classes have been selected for training and testing purposes (see Table 9). We have used PFS with DT as the inner classifier and followed the same setup to compare PFS-DT with the method proposed in (Ghaddar & Naoum-Sawaya, 2018). To get unbiased results, we run PFS-DT 10 times where each time we shuffled and constructed test and train datasets based on the configuration in Table 9.…”

Section: A Quantified Measurementioning

confidence: 99%

See 1 more Smart Citation

A Feature Selection based on perturbation theory

Anaraki

Usefi

2019

Expert Systems with Applications

View full text Add to dashboard Cite

Consider a supervised dataset D = [A | b], where b is the outcome column, rows of D correspond to observations, and columns of A are the features of the dataset. A central problem in machine learning and pattern recognition is to select the most important features from D to be able to predict the outcome.In this paper, we provide a new feature selection method where we use perturbation theory to detect correlations between features. We solve AX = b using the method of least squares and singular value decomposition of A. In practical applications, such as in bioinformatics, the number of rows of A (observations) are much less than the number of columns of A (features). So we are dealing with singular matrices with big condition numbers. Although it is known that the solutions of least square problems in singular case are very sensitive to perturbations in A, our novel approach in this paper is to prove that the correlations between features can be detected by applying perturbations to A.The effectiveness of our method is verified by performing a series of comparisons with conventional and novel feature selection methods in the literature.It is demonstrated that in most situations, our method chooses considerably less number of features while attaining or exceeding the accuracy of the other methods.

show abstract

Section: A Quantified Measurementioning

confidence: 99%

“…In order to find the highest classification accuracy, the authors in (Ghaddar & Naoum-Sawaya, 2018) have applied their method FS-SVM and limited the selected subset of features to range from 2% to 20% of total number of features. In turn, the running time of FS-SVM is very high.…”

Section: A Quantified Measurementioning

confidence: 99%

A Feature Selection based on perturbation theory

Anaraki

Usefi

2019

Expert Systems with Applications

View full text Add to dashboard Cite

show abstract

“…Dengan jumlah fitur yang sangat tinggi, SVM telah menunjukkan hasil yang baik. Pendekatan klasifikasi dan seleksi fitur yang diajukan, sederhana, dan alur yang dapat ditelusuri, dan mencapai rata-rata error yang rendah [17]. Dalam penelitian yang dilakukan oleh Xuchan Ju dkk.…”

Section: B Support Vector Machinesunclassified

“…Khusus metode SVM, penggunaan proses peringkasan tidak meningkatkan performansinya. Hasil yang diperoleh ini berbeda dengan yang didapatkan oleh Ghaddar, B., & Naoum-Sawaya, J. yang menunjukkan performasi SVM semakin baik dengan adanya proses seleksi terlebih dahulu pada kasus analisis sentiment [17]. Nilai akurasi SVM yang justru berkurang setelah dilakukan proses peringkasan memunculkan dugaan bahwa seleksi fitur dengan peringkasan justru telah menghilangkan fitur kata yang relevan dalam menentukan kelas dari data yang diuji.…”

Section: Pembahasanunclassified

Peringkasan dan Support Vector Machine pada Klasifikasi Dokumen

Widiastuti

Rainarli²,

Dewi³

2017

INFOTEL

View full text Add to dashboard Cite

Abstrak -Klasifikasi adalah proses pengelompokkan objek yang memiliki karakteristik atau ciri yang sama ke dalam beberapa kelas. Klasifikasi dokumen secara otomatis dapat dilakukan dengan menggunakan ciri atau fitur kata yang muncul pada dokumen latih. Jumlah dokumen yang besar dan banyak mengakibatkan jumlah kata yang muncul sebagai fitur akan bertambah. Oleh karena itu, peringkasan dipilih untuk mereduksi jumlah kata yang digunakan dalam proses klasifikasi. Untuk proses klasifikasi digunakan metode Support Vector Machine (SVM) untuk multikelas. SVM dipilih karena dianggap memiliki reputasi yang baik dalam klasifikasi. Penelitian ini menguji penggunaan ringkasan sebagai seleksi fitur dalam klasifikasi dokumen. Peringkasan menggunakan kompresi 50 %. Hasil yang diperoleh menunjukkan bahwa proses peringkasan tidak mempengaruhi nilai akurasi dari klasifikasi dokumen yang menggunakan SVM.Akan tetapi, penggunaan peringkasan berpengaruh pada peningkatan hasil akurasi dari metode klasifikasiSimple Logistic Classifier (SLC). Hasil pengujian metode klasifikasi menunjukkan bahwa penggunaan metode Naïve Bayes Multinomial (NBM) menghasilkan akurasi yang lebih baik dari pada metode SVM.Kata kunci -peringkasan, klasifikasi, SVM, seleksi fitur, dokumen Abstract -Classification is the process of grouping objects that have the same features or characteristics into several classes. The automatic documents classification use words frequency that appears on training data as features.The large number of documents cause the number of words that appears as a feature will increase.Therefore, summaries are chosen to reduce the number of words that used in classification.The classification uses multiclass Support Vector Machine (SVM) method.SVM was considered to have a good reputation in the classification. This research tests the effect of summary as selection features into documents classification.The summariesreduce text into 50%.A result obtained that the summaries did not affect value accuracy of classification of documents that use SVM.But, summaries improve the accuracy of Simple Logistic Classifier. The classification testing shows that the accuracy of Naïve Bayes Multinomial (NBM) better than SVM.

show abstract

“…Although GP algorithms have been used to evolve probabilistic trees that search for the optimal topology in bioinformatics (Won et al, 2007) and stock trading (Chen et al, 2009;Ghaddar et al, 2016), to the best of our knowledge, this is the first work that a MOGP algorithm has been used as a multi-class classifier to construct a classification-HMM hybrid model for solving sequential learning problems. Our model can be of interest and easily adapted to other relevant domains in business analytics, such as consumer choice modelling (Sandkci et al, 2008;Blanchet et al, 2016) and high dimensional business data classification or dimension reduction (Debaere et al, 2018;Ghaddar & Naoum-Sawaya, 2018).…”

Section: Introductionmentioning

confidence: 99%

A hybrid model for predicting human physical activity status from lifelogging data

Chen

Allinson

et al. 2020

European Journal of Operational Research

View full text Add to dashboard Cite

One trend in the recent healthcare transformations is people are encouraged to monitor and manage their health based on their daily diets and physical activity habits. However, much attention of the use of operational research and analytical models in healthcare has been paid to the systematic level such as country or regional policy making or organisational issues. This paper proposes a model concerned with healthcare analytics at the individual level, which can predict human physical activity status from sequential lifelogging data collected from wearable sensors. The model has a two-stage hybrid structure (in short, MOGP-HMM) -a multi-objective genetic programming (MOGP) algorithm in the first stage to reduce the dimensions of lifelogging data and a hidden Markov model (HMM) in the second stage for activity status prediction over time. It can be used as a decision support tool to provide real-time monitoring, statistical analysis and personalized advice to individuals, encouraging positive attitudes towards healthy lifestyles. We validate the model with the real data collected from a group of participants in the UK, and compare it with other popular two-stage hybrid models. Our experimental results show that the MOGP-HMM can achieve comparable performance. To the best of our knowledge, this is the very first study that uses the MOGP in the hybrid two-stage structure for individuals' activity status prediction. It fits seamlessly with the current trend in the UK healthcare transformation of patient empowerment as well as contributing to a strategic development for more efficient and cost-effective provision of healthcare.

show abstract

High dimensional data classification and feature selection using support vector machines

Cited by 218 publications

References 22 publications

A Feature Selection based on perturbation theory

A Feature Selection based on perturbation theory

Peringkasan dan Support Vector Machine pada Klasifikasi Dokumen

A hybrid model for predicting human physical activity status from lifelogging data

Contact Info

Product

Resources

About