A membership based neutrosophic approach for supervised fingerprint image classification

Abstract

The Neutrosophic Sets (NS) mathematical model is a sophisticated paradigm that effectively addresses uncertainty. This article provides four different methods for the extraction of visual features. The proposal has been investigated with regard to both neutrosophic sets and single-valued NS. The article primarily examines two distinct features: binary and self-intensity approaches. Following that, an attempt was made to classify the images using machine learning techniques. The main objective of this article was exclusively on supervised classification algorithms. The classification of images was performed by using Decision Tree (DT), Random Forest (RF), K Nearest Neighbour (KNN),Naive Bayes (NB), and Logistic Regression (LR) algorithms. Since we have an interest in biometric images, the fingerprint image dataset was chosen for classification. The methods proposed in that research are known to as Membership Based Neutrosophic Binary Image (MBNIB), Membership Based Neutrosophic Self Intensity Image (MBNISI ), Membership Based Single Valued Neutrosophic Binary Image (MBSV NIB), and Membership Based Single Valued Neutrosophic Self Intensity Image (MBSV NISI ). The proposal possesses a range of improvement accuracy ranging from 5% to 58%.

Keywords

• machine learning
• decision tree
• random forest
• naive bayes
• logistic regression
• neutrosophic image
• fingerprint image