AL-Qadisiyah Journal For Administrative and Economic sciences

Proposing a Robust Discriminant Analysis Method with Application to Genetic Sequence Classification Using GenBank Data

Document Type : Research Paper

Authors

Mohammed H. AL-Sharoot
Souada bdel husseinmusa

University of AL-Qadisiya

https://doi.org/10.33916/qjae.2026.01134139
Abstract
This research proposes a robust discriminant analysis method for classifying genetic sequences related to breast cancer, using real genomic data obtained from GenBank. Traditional discriminant methods such as Fisher’s Linear Discriminant Function and Sherrod’s model are known to be sensitive to outliers, which are common in genomic datasets due to biological variability and sequencing errors. To overcome this, the proposed method integrates robust estimators of location and scale along with a novel reweighting algorithm that reduces the influence of outlying gene expressions. The model’s performance is evaluated through a simulation study under clean and contaminated conditions, showing improved classification accuracy and reduced misclassification rates. For real data application, genetic sequences of breast tissue samples labeled as tumor and normal were analyzed. The robust model achieved superior accuracy in separating cancerous from non-cancerous samples, confirming its practical value in biomedical classification tasks involving noisy or high-variability data.

Keywords

Robust Discriminant Analysis
Breast Cancer
Genetic Sequences
Outliers
GenBank
Tumor Classification
Crossmark
AL-Qadisiyah Journal For Administrative and Economic sciences
Volume 28, Issue 1
Spring 2026
Pages 134-139

Home

Guide for Authors

Submit Manuscript

Reviewers

Contact Us