Integrating Traditional Diagnostics Methods Through  Neutrosophic Set Based Cancer Analysis Using Machine  Learning Techniques

P. Tharaniya; M. Raji; R. Rajalakshmi; Surapati Pramanik; R. Balapriya; M. Gayathri Lakshmi

Authors

P. Tharaniya Department of Mathematics, Chennai Institute of Technology, Kundrathur, Chennai – 600 069, Tamilnadu, India,
M. Raji Department of Mathematics, Vels Institute of Science, Technology and Advanced Studies, Pallavaram, Chennai-600 117, Tamilnadu, India,
R. Rajalakshmi Department of Mathematics, Panimalar Engineering College, Chennai – 600 123, Tamilnadu, India,
Surapati Pramanik Department of Mathematics, Nandalal Ghosh B.T. College, Panpur, Narayanpur, Dist- North 24 Parganas, West Bengal, India, PIN-743126,
R. Balapriya Department of Mathematics, Saveetha School of Engineering, Saveetha Institute of Medical and Technical Sciences, Thandalam, Chennai – 602 105, Tamilnadu, India,
M. Gayathri Lakshmi Department of Mathematics, Saveetha Engineering College, Chennai – 602 105, Tamilnadu, India,

Keywords:

Neutrosophic sets; machine learning; neutrosophic confusion matrix; image processing

Abstract

This paper introduces a novel approach for utilizing Neutrosophic Sets through machine
learning to enhance classification accuracy by incorporating truth, falsity, and indeterminacy in
decision-making. Traditional cancer cell classification models struggle to handle indeterminate
cases effectively. A Neutrosophic Confusion Matrix (NCM) is developed to extend conventional
performance evaluation metrics, considering the probability distribution of positive, negative, and
neutral classifications. This framework enables a more comprehensive assessment of classification
reliability, particularly in ambiguous cases where traditional machine-learning models exhibit
limitations. The proposed approach for classification uses two significant imagery features: texture
contrast and color saturation as well as neutrosophic sets that can effectively differentiate between
benign, malignant, and healthy skin lesions through Machine Learning concepts. Through empirical
validation of medical datasets, this work establishes Neutrosophic based classification as a powerful
tool for improving the accuracy and robustness of cancer diagnosis.

DOI: 10.5281/zenodo.16897803