Scholarlink

This paper explores the classification of DNA sequencesDNA sequences through a multi-step approach that integrates sequence encoding, dimensionality reduction techniques (principal component analysis (PCA))Principal Component Analysis (PCA), t-distributed Stochastic neighbour embeddingStochastic Neighbour Embedding (t-SNE), uniform manifold approximationManifold Approximation and projection (UMAP), and a random forestRandom Forest classifier for predictive analysis. The aim is to analyse the relationships between these dimensionality reduction techniques, visualize the separability of different DNA sequence classes, and assess classification performance using standard metrics such as accuracy, precision, F1 score, and AUC. Additionally, the paper examines correlations between the reduced feature spaces generated by PCAPrincipal Component Analysis (PCA), t-SNE, and UMAP to evaluate their similarities and differences in the context of multi-class classification. This study demonstrates the effectiveness of combining dimensionality reduction with machine learningMachine learning for DNA sequence analysis and classification.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Analysing and Classifying DNA Sequences Using Dimensionality Reduction Techniques and Random Forests

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Elias Tabane,
  • Ernest Mnkandla,
  • Zenghui Wang

摘要

This paper explores the classification of DNA sequencesDNA sequences through a multi-step approach that integrates sequence encoding, dimensionality reduction techniques (principal component analysis (PCA))Principal Component Analysis (PCA), t-distributed Stochastic neighbour embeddingStochastic Neighbour Embedding (t-SNE), uniform manifold approximationManifold Approximation and projection (UMAP), and a random forestRandom Forest classifier for predictive analysis. The aim is to analyse the relationships between these dimensionality reduction techniques, visualize the separability of different DNA sequence classes, and assess classification performance using standard metrics such as accuracy, precision, F1 score, and AUC. Additionally, the paper examines correlations between the reduced feature spaces generated by PCAPrincipal Component Analysis (PCA), t-SNE, and UMAP to evaluate their similarities and differences in the context of multi-class classification. This study demonstrates the effectiveness of combining dimensionality reduction with machine learningMachine learning for DNA sequence analysis and classification.