International Journal of Computing and Digital Systems ISSN (2210-142X) Int. J. Com. Dig. Sys. 16, No.1 (Sep-24) http://dx.doi.org/10.12785/ijcds/160197 Online Signature Classification Based on Dynamic Nature of Features Selection Framework Akhilesh Kumar Singh 1 , Surabhi Kesarwani 2 , Anushree 3 , Pawan Kumar Verma 1,* , Nitin Rakesh 4 and Monali Gulhane 4 1 Sharda University, Greater Noida Uttar Pradesh, India 2 Greater Noida Institute of Technology (Engineering Institute), Uttar Pradesh, India 3 GLA University, Mathura, Uttar Pradesh, India 4 Symbiosis Institute of Technology, Nagpur Campus, Symbiosis International (Deemed University), Pune, Maharashtra, India Received 15 Mar. 2024, Revised 27 May 2024, Accepted 31 May 2024, Published 15 Sep. 2024 Abstract: In the recent digital age, online signature verification plays a crucial role in authentication, including security standards across many industries, such as financial, legal, and e-commerce. Bank’srld Bank’s data shows the global digital economy is growing fast, with internet usage for nearly 60% of people worldwide. According to numbers from the International Telecommunications Union, over 4.7 billion individuals have become internet users. With so much internet online, security and trust for online transactions are essential issues. Forensics and biometrics are emerging as key players in this area. Verifying signatures digitally is one important use. As in the study mentioned earlier, using machine learning can help make signature verification systems more accurate and reliable. Our study describes an online verification method using machine learning based on a signature’s dynamic features and compares the outcomes to methods already in use. The online signature verification has been validated using supervised learning (K-nearest neighbor (KNN)). This research aimed to enhance authenticity and reduce the occurrence of false positives as its primary objectives. The outcomes show that this methodology has better authenticity than the current methods. The Signature Verification System (SVS) 2004-based signature datasets are utilized in the tests. Keywords: Online Signature Verification, Signature features, KNN (k Nearest Neighbor), Machine Learning 1. INTRODUCTION Biometric systems have arisen as innovative security solutions in pattern recognition and E-systems, thanks to the rapid progress of information technology. Physiological and behavioral biometric systems are the two primary categories of biometric systems [1]. Physiological features are distinct human body properties that are static [2], [3]. Behavioral characteristics, on the other hand, are fluid and can change over time depending on mood, age, and other circum- stances. Behavioral qualities are influenced by gait, signa- ture, handwriting, voice, keyboard, and other modalities. A handwritten signature is widely accepted by institutional and financial institutions as a reliable method of personal recognition [4], [5]. The commercial and banking sectors, as well as many other businesses, are now quickly utilizing digital signature systems treated specifically to permit pur- chases and transfers. Signatures represent human biometrics that can vary because of certain conditions, such as age, mood, and climate, so two individual signatures cannot fit each other exactly [6]. The Signature Verification System, also called SVS, recognizes and validates a handwritten signature of authenticity. Static (offline) and dynamic SVS are two types of SVS (online). User signatures are digitized using a scanner or a camera from paper in an offline system, whereas they are digitized using a scanner or a camera in an online system [7]. The stability of dynamic features is minimized in most existing online handwritten signature authentication systems since they compare different signa- tures using the same homogenous feature sets for different nonidentical users [5], [8]. Enrollment and verification phases are typical in on- line signature verification systems. Users supply their self- reference-based signatures during the enrolling process, which are then included in a system that makes use of feature extraction methods. The system compares and analo- gizes a signature query onto reference-dependent signatures and applies matching algorithms to approve or repudiate it during the verification phase [9]. The signature version system’s efficiency can be improved by focusing on fea- ture extraction and classification methodologies. In online- Email: akhilesh.singh1@sharda.ac.in, surabhikesarwani.mca@gniot.net.in, anushree.gla@gla.ac.in, abes.pawan@gmail.com,nitin.rakesh@gmail.com,monaligulhane4@gmail.com https:// journal.uob.edu.bh