An efficient feature selection framework based on information theory for high dimensional data

Manikandan, G.; Abirami, S.

doi:10.1016/j.asoc.2021.107729

Cited by 32 publications

(11 citation statements)

References 58 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…MCTS is a relatively new technique and is being deployed in various application domains such as games [14], feature selection [30] and parameter tuning [17]. MCTS has rarely been used for combinatorial optimization problems.…”

Section: Related Workmentioning

confidence: 99%

A 2-Stage Approach for the Nurse Rostering Problem

et al. 2022

View full text Add to dashboard Cite

In this paper, we are addressing the NP-hard nurse rostering problem utilizing a 2-stage approach. In stage one, Monte Carlo Tree Search (MCTS) and Hill Climbing (HC) are hybridized in finding a feasible solution (satisfying all the hard constraints). We propose a new constant C value (which balances search diversification and intensification of MCTS) and tree policy/node selection function in the selection procedure of MCTS. In stage two, the feasible solution is further improved using Iterated Local Search (ILS) with Variable Neighbourhood Descent as the local search component. We introduce several unique neighbourhood structures for the ILS. In addition, we propose a novel perturbation strategy to allow the search to escape from local optimum. The proposed methodology is tested on the Shift Scheduling dataset (24 benchmark instances). New best results are reported for seven and two instances for the 10 and 60 minutes run respectively. An in-depth discussion on the attributes of the proposed methodology that lead to its good performance is provided.INDEX TERMS nurse rostering, hill climbing, monte carlo tree search, iterated local search, variable neighbourhood descent.

show abstract

Section: Related Workmentioning

confidence: 99%

A 2-Stage Approach for the Nurse Rostering Problem

et al. 2022

View full text Add to dashboard Cite

show abstract

“…Walaupun demikian algoritma pohon keputusan C4.5 ini memiliki kelemahan pada data tidak seimbang dan berdimensi tinggi yang memiliki fitur banyak sehingga dapat menurunkan kinerja dari algoritma ini [9]. Untuk mengurai fitur fitur penting dan relevan pemilihan fitur adalah cara yang bisa dipilih dalam meningkatkan kinerja algoritma pohon keputusan C4.5 [10].…”

Section: A Pendahuluanunclassified

Meningkatkan Kinerja Decision Tree C4.5 dengan Seleksi Fitur Korelasi Pearson pada Deteksi Penyakit Diabetes

Hanif

Setiaji

2022

ijcs

View full text Add to dashboard Cite

Diabetes sebuah penyakit yang menjadi momok seluruh dunia. Kerugianya tidak hanya pada penderita sendiri tetapi juga merambah ke banyak sektor. Baik di sektor pelayanan kesehatan dan sektor financial yang sangat menjadi beban tinggi yang perlu ditangani dengan baik dengan jalan pendeteksian penyakit diabetes sejak dini. Salah satu pendeteksian dini penyakit diabetes dapat memanfaatkan algoritma machine learning pada bidang data mining. Algoritma C4.5 merupakan algoritma machine learning yang memiliki tingkat akurasi dan kecepatan perhitungan tinggi dalam klasifikasi. Namun demikian algoritma C4.5 terganggu dengan data tak seimbang dan fitur data berdimensi tinggi. Pemanfaatan seleksi fitur menjadi salah satu penyelesain masalah data berdimensi tinggi. Algoritma Korelasi Pearson memiliki kemampuan dalam mengukur informasi antar fitur dan diterapkan dalam penelitian ini. Penggunaan Korelasi Pearson dianggap berhasil dalam meningkatkan kinerja algoritma C4.5 dalam deteksi awal penyakit diabetes. Keberhasilan ini terlihat pada hasil akurasi sebesar 95.31% tanpa korelasi pearson menjadi 96.16% dengan pemanfaatan korelasi pearson.

show abstract

“…Yu et al 11 used the Fisher score evaluation criterion to carry out feature selection and eliminate redundant features to complete mechanical fault diagnosis. Manikan-dan et al 12 proposed a feature selection technique based on mutual information and Monte Carlo to remove redundant features. The evaluation criteria should be taken into comprehensive consideration to select the reasonable degradation index, which is sensitive to degradation trends and has sound robustness, correlation, and monotonicity.…”

Section: Introductionmentioning

confidence: 99%

Mechanical Equipment Performance Degradation Assessment Method Based on Improved Degradation Feature Extraction and Sample Completeness

Lv¹,

Xiong²,

Wang³

2023

The International Journal of Acoustics and Vibration

View full text Add to dashboard Cite

Performance degradation assessment (PDA), as an important part of health management, is playing a crucial role in evaluating the degenerate state of mechanical equipment. To evaluate the performance degradation state more effectively by using monitoring data of mechanical equipment, a PDA method based on improved degradation feature extraction technology and considering sample completeness is proposed in this paper. Firstly, multiple degradation features, including statistical features and intrinsic energy features, are extracted to construct a high-dimensional feature set calculated by time-domain analysis and the EMD method. Then, a sensitive feature set, which has significant robustness, correlation, and monotonicity in the degenerate process, is reduced and processed by the improved PCA method to obtain the final health index. Secondly, the degradation assessment model is established by considering the completeness of the samples. As for the complete sample dataset, which has both normal and failure state data, a logistic regression model (LRM) is built to assess the performance degradation status. In addition, an SVDD model is established for the incomplete sample dataset, which only has normal state data. Finally, the reliability of the proposed method is verified by the public XJTU-SY dataset, and the comparative experiment further verifies the superiority of this method.

show abstract

An efficient feature selection framework based on information theory for high dimensional data

Cited by 32 publications

References 58 publications

A 2-Stage Approach for the Nurse Rostering Problem

A 2-Stage Approach for the Nurse Rostering Problem

Meningkatkan Kinerja Decision Tree C4.5 dengan Seleksi Fitur Korelasi Pearson pada Deteksi Penyakit Diabetes

Mechanical Equipment Performance Degradation Assessment Method Based on Improved Degradation Feature Extraction and Sample Completeness

Contact Info

Product

Resources

About