On Combining Feature Selection and Over-Sampling Techniques for Breast Cancer Prediction

Huang, Min; Chiu, Chien Hung; Tsai, Chih‐Fong; Lin, Wei

doi:10.3390/app11146574

Cited by 17 publications

(5 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…These results agree with other authors who studied the combined feature selection and resampling method for imbalance data learning and found that in 79% of the study cases, balancing before feature selection improves the results [ 59 ]. We also showed the feasibility of combining both the feature selection and resampling methods in the same iterative process to deal with imbalanced data, in contrast to resampling before or after feature selection [ 59 , 60 ]. Balancing validation data to deal with the imbalance data problem was similar to the strategy proposed by Jain et al, who used a weighted sum of recall and specificity as the fitness function [ 61 ].…”

Section: Discussionmentioning

confidence: 99%

Combination of Feature Selection and Resampling Methods to Predict Preterm Birth Based on Electrohysterographic Signals from Imbalance Data

Nieto-del-Amor

Prats-Boluda

Garcia-Casado

et al. 2022

Sensors

View full text Add to dashboard Cite

Due to its high sensitivity, electrohysterography (EHG) has emerged as an alternative technique for predicting preterm labor. The main obstacle in designing preterm labor prediction models is the inherent preterm/term imbalance ratio, which can give rise to relatively low performance. Numerous studies obtained promising preterm labor prediction results using the synthetic minority oversampling technique. However, these studies generally overestimate mathematical models’ real generalization capacity by generating synthetic data before splitting the dataset, leaking information between the training and testing partitions and thus reducing the complexity of the classification task. In this work, we analyzed the effect of combining feature selection and resampling methods to overcome the class imbalance problem for predicting preterm labor by EHG. We assessed undersampling, oversampling, and hybrid methods applied to the training and validation dataset during feature selection by genetic algorithm, and analyzed the resampling effect on training data after obtaining the optimized feature subset. The best strategy consisted of undersampling the majority class of the validation dataset to 1:1 during feature selection, without subsequent resampling of the training data, achieving an AUC of 94.5 ± 4.6%, average precision of 84.5 ± 11.7%, maximum F1-score of 79.6 ± 13.8%, and recall of 89.8 ± 12.1%. Our results outperformed the techniques currently used in clinical practice, suggesting the EHG could be used to predict preterm labor in clinics.

show abstract

Section: Discussionmentioning

confidence: 99%

Combination of Feature Selection and Resampling Methods to Predict Preterm Birth Based on Electrohysterographic Signals from Imbalance Data

Nieto-del-Amor

Prats-Boluda

Garcia-Casado

et al. 2022

Sensors

View full text Add to dashboard Cite

show abstract

“…Furthermore, the paper does not extensively compare the proposed framework with more recent and state-of-the-art clustering methods. Huang et al [9], proposed a methodology to improve the accuracy of breast cancer prediction models. The paper highlights the challenges posed by imbalanced datasets in breast cancer prediction, where the minority class (cancer cases) has significantly fewer samples than the majority class (non-cancer cases).…”

Section: Figure 1 Progressive Learning Curvementioning

confidence: 99%

Improving time efficiency in big data through progressive sampling-based classification model

Bangera,

Kayarvizhy,

Luharuka

et al. 2024

IJEECS

View full text Add to dashboard Cite

<span>The proposed system aims to overcome challenges posed by large databases, data imbalance, heterogeneity, and multidimensionality through progressive sampling as a novel classification model. It leverages sampling techniques to enhance processing performance and overcome memory restrictions. The random forest regressor feature importance technique with the gini significance method is employed to identify important characteristics, reducing the data’s features for classification. The system utilizes diverse classifiers such as random forest, ensemble learning, support vector machine (SVM), k-nearest neighbors’ algorithm (KNN), and logistic regression, allowing flexibility in handling different data types and achieving high accuracy in classification tasks. By iteratively applying progressive sampling to the dataset with the best features, the proposed technique aims to significantly improve performance compared to using the entire dataset. This approach focuses computational resources on the most informative subsets of data, reducing time complexity. Results show that the system can achieve over 85% accuracy even with only 5-10% of the original data size, providing accurate predictions while reducing data processing requirements. In conclusion, the proposed system combines progressive sampling, feature selection using random forest regressor feature importance (RFRFI-PS), and a range of classifiers to address challenges in large databases and improve classification accuracy. It demonstrates promising results in accuracy and time complexity reduction.</span>

show abstract

“…Dengan melakukan seleksi fitur terlebih dahulu, jumlah fitur akan dikurangi sehingga SMOTE dapat berfungsi lebih efektif dalam menghasilkan sampel sintetis yang merepresentasikan kelas minoritas secara akurat. Hal tersebut juga disarankan oleh peneliti terdahulu (Huang et al, 2021).…”

Section: Smoteunclassified

Seleksi Fitur dengan Particle Swarm Optimization pada Klasifikasi Penyakit Parkinson Menggunakan XGBoost

Kurnia,

Itqan Mazdadi,

Kartini

et al. 2023

JTIIK

View full text Add to dashboard Cite

Penyakit Parkinson merupakan gangguan pada sistem saraf pusat yang mempengaruhi sistem motorik. Diagnosis penyakit ini cukup sulit dilakukan karena gejalanya yang serupa dengan penyakit lain. Saat ini diagnosa dapat dilakukan menggunakan machine learning dengan memanfaatkan rekaman suara pasien. Fitur yang dihasilkan dari ekstraksi rekaman suara tersebut relatif cukup banyak sehingga seleksi fitur perlu dilakukan untuk menghindari memburuknya kinerja sebuah model. Pada penelitian ini, Particle Swarm Optimization digunakan sebagai seleksi fitur, sedangkan XGBoost akan digunakan sebagai model klasifikasi. Selain itu model juga akan diterapkan SMOTE untuk mengatasi masalah ketidakseimbangan kelas data dan hyperparameter tuning pada XGBoost untuk mendapatkan hyperparameter yang optimal. Hasil pengujian menunjukkan bahwa nilai AUC pada model dengan seleksi fitur tanpa SMOTE dan hyperparameter tuning adalah 0,9325, sedangkan pada model tanpa seleksi fitur hanya mendapat nilai AUC sebesar 0,9250. Namun, ketika kedua teknik SMOTE dan hyperparameter tuning digunakan bersamaan, penggunaan seleksi fitur mampu memberikan peningkatan kinerja pada model. Model dengan seleksi fitur mendapat nilai AUC sebesar 0,9483, sedangkan model tanpa seleksi fitur hanya mendapat nilai AUC sebesar 0,9366. Abstract Parkinson's disease is a disorder of the central nervous system that affects the motor system. Diagnosis of this disease is quite difficult because the symptoms are similar to other diseases. Currently, diagnosis can be done using machine learning by utilizing patient voice recordings. The features generated from the extraction of voice recordings are relatively large, so feature selection needs to be done to avoid deteriorating the performance of a model. In this research, Particle Swarm Optimization is used as feature selection, while XGBoost will be used as a classification model. In addition, the model will also be applied SMOTE to overcome the problem of data class imbalance and hyperparameter tuning on XGBoost to get optimal hyperparameters. The test results show that the AUC value on the model with feature selection without SMOTE and hyperparameter tuning is 0.9325, while the model without feature selection only gets an AUC value of 0.9250. However, when both SMOTE and hyperparameter tuning techniques are used together, the use of feature selection is able to provide improved performance on the model. The model with feature selection gets an AUC value of 0.9483, while the model without feature selection only gets an AUC value of 0.9366.

show abstract

On Combining Feature Selection and Over-Sampling Techniques for Breast Cancer Prediction

Cited by 17 publications

References 27 publications

Combination of Feature Selection and Resampling Methods to Predict Preterm Birth Based on Electrohysterographic Signals from Imbalance Data

Combination of Feature Selection and Resampling Methods to Predict Preterm Birth Based on Electrohysterographic Signals from Imbalance Data

Improving time efficiency in big data through progressive sampling-based classification model

Seleksi Fitur dengan Particle Swarm Optimization pada Klasifikasi Penyakit Parkinson Menggunakan XGBoost

Contact Info

Product

Resources

About