Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Feature Selection via Robust Weighted Score for High Dimensional Binary Class-Imbalanced Gene Expression Data

Zardad Khan, Amjad Ali, Saeed Aldahmani

TL;DR

The paper tackles binary gene expression classification under extreme class imbalance by introducing ROWSU, a three-stage feature selection framework that balances data via minority-mean augmentation, selects a minimal feature subset with a greedy mask-based search, and refines features through a robust Fisher score multiplied by SVM-derived feature weights. The final feature set combines the greedy subset with top ROW features, producing a discriminative gene panel suitable for high-dimensional data. Across six benchmark datasets and two classifiers, ROWSU yields superior accuracy and sensitivity relative to Fisher, Wilcoxon, SNR, POS, and MRMR, with results visualized through stability and boxplots. This approach offers a practical and effective solution for robust gene selection in imbalanced genomic datasets, facilitating more reliable downstream classification.

Abstract

In this paper, a robust weighted score for unbalanced data (ROWSU) is proposed for selecting the most discriminative feature for high dimensional gene expression binary classification with class-imbalance problem. The method addresses one of the most challenging problems of highly skewed class distributions in gene expression datasets that adversely affect the performance of classification algorithms. First, the training dataset is balanced by synthetically generating data points from minority class observations. Second, a minimum subset of genes is selected using a greedy search approach. Third, a novel weighted robust score, where the weights are computed by support vectors, is introduced to obtain a refined set of genes. The highest-scoring genes based on this approach are combined with the minimum subset of genes selected by the greedy search approach to form the final set of genes. The novel method ensures the selection of the most discriminative genes, even in the presence of skewed class distribution, thus improving the performance of the classifiers. The performance of the proposed ROWSU method is evaluated on $6$ gene expression datasets. Classification accuracy and sensitivity are used as performance metrics to compare the proposed ROWSU algorithm with several other state-of-the-art methods. Boxplots and stability plots are also constructed for a better understanding of the results. The results show that the proposed method outperforms the existing feature selection procedures based on classification performance from k nearest neighbours (kNN) and random forest (RF) classifiers.

Feature Selection via Robust Weighted Score for High Dimensional Binary Class-Imbalanced Gene Expression Data

TL;DR

The paper tackles binary gene expression classification under extreme class imbalance by introducing ROWSU, a three-stage feature selection framework that balances data via minority-mean augmentation, selects a minimal feature subset with a greedy mask-based search, and refines features through a robust Fisher score multiplied by SVM-derived feature weights. The final feature set combines the greedy subset with top ROW features, producing a discriminative gene panel suitable for high-dimensional data. Across six benchmark datasets and two classifiers, ROWSU yields superior accuracy and sensitivity relative to Fisher, Wilcoxon, SNR, POS, and MRMR, with results visualized through stability and boxplots. This approach offers a practical and effective solution for robust gene selection in imbalanced genomic datasets, facilitating more reliable downstream classification.

Abstract

In this paper, a robust weighted score for unbalanced data (ROWSU) is proposed for selecting the most discriminative feature for high dimensional gene expression binary classification with class-imbalance problem. The method addresses one of the most challenging problems of highly skewed class distributions in gene expression datasets that adversely affect the performance of classification algorithms. First, the training dataset is balanced by synthetically generating data points from minority class observations. Second, a minimum subset of genes is selected using a greedy search approach. Third, a novel weighted robust score, where the weights are computed by support vectors, is introduced to obtain a refined set of genes. The highest-scoring genes based on this approach are combined with the minimum subset of genes selected by the greedy search approach to form the final set of genes. The novel method ensures the selection of the most discriminative genes, even in the presence of skewed class distribution, thus improving the performance of the classifiers. The performance of the proposed ROWSU method is evaluated on gene expression datasets. Classification accuracy and sensitivity are used as performance metrics to compare the proposed ROWSU algorithm with several other state-of-the-art methods. Boxplots and stability plots are also constructed for a better understanding of the results. The results show that the proposed method outperforms the existing feature selection procedures based on classification performance from k nearest neighbours (kNN) and random forest (RF) classifiers.
Paper Structure (15 sections, 21 equations, 13 figures, 7 tables, 1 algorithm)

This paper contains 15 sections, 21 equations, 13 figures, 7 tables, 1 algorithm.

Table of Contents

  1. Introduction
  2. Related work
  3. The Robust Weighted Score for Unbalanced Data (ROWSU)
  4. Balancing The Data
  5. Selecting Minimum Subset of Features
  6. Robust Fisher Score
  7. Feature Weights ($w$)
  8. The Robust Weighted Score
  9. Final Set of the Most Discriminative Features
  10. Algorithm
  11. Experiments and results
  12. Benchmark datasets
  13. Experimental setup
  14. Results
  15. Conclusion

Figures (13)

  • Figure 1: Flowchart of the proposed ROWSU algorithm.
  • Figure 2: Matplots of the results computed for the top $p^*$ features by the proposed method and other state-of-the-art procedures using the $D_1$ dataset.
  • Figure 3: Matplots of the results computed for the top $p^*$ features by the proposed method and other state-of-the-art procedures using the $D_2$ dataset.
  • Figure 4: Matplots of the results computed for the top $p^*$ features by the proposed method and other state-of-the-art procedures using the $D_3$ dataset.
  • Figure 5: Matplots of the results computed for the top $p^*$ features by the proposed method and other state-of-the-art procedures using the $D_4$ dataset.
  • ...and 8 more figures