A Tangent Similarity-Based Pentapartitioned Neutrosophic Soft Topological Framework for Cancer Diagnosis

Abstract

This research paper introduces the concept of pentapartitioned neutrosophic soft topological spaces (PNSTS). In PNSTS, the indeterminacy is divided into three components: truth (ReT), hesitation (H), and falsehood (ReF). Fundamental topological concepts and tangent-based similarity measures have been discussed and applied to a cancer data set to ascertain the similarity between patients and potential treatments. To analyze results more effectively, we have employed Heatmaps, 3D surface plotting, PCA, t-SNE, and DBSCAN clustering. The experimental results demonstrated that the proposed framework was highly effective, as the maximum similarity value reached 0.7508. This indicates that the patient’s traits and the proposed diagnosis clearly aligned. This demonstrates that the proposed method can be effectively employed for intelligent medical decision-making.