International Journal of Computer and Information Technology(2279-0764)

Identification of Medical Mask Use by Applying the Convolutional Neural Network Algorithm and the Gabor Filter with Multiclass Classification

Muh. Arifandi, Erik Iman Heri Ujianto — Sat, 30 Sep 2023 00:00:00 +0000

Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-COV-2) causes global pandemics and makes countries around the world lock down fortourists. This action is required to prevent the spread of viruses that take 14 days to disappear. SARS-COV-2 can easily infect individuals through a droplet. Thus, the governments of every country worldwide recommend wearing medical masks to prevent the spread of viruses, as well as maintaining distance during activities with others and washing hands frequently. Medical masks become efficient if their application is precise, owing to a lack of knowledge and self-awareness to preserve their distance and wash their hands. This paper proposes a Convolutional Neural Network (CNN) with Gabor filter implementation. The simulation uses a mask on a dataset with over 70,000 individual photos. The results demonstrated that the proposed CNN-Gabor model in this work could effectively classify the position of the mask when compared to the CNN model without the Gabor filter.

Fraud Detection in Motor Insurance Claims Using Supervised Learning Techniques: A Review

David Gichohi Maina, Juliet Chebet Moso, Patrick Kinyua Gikunda — Sat, 30 Sep 2023 00:00:00 +0000

Fraudulent claims have been a big drawback in motor insurance despite the insurance industry having vast amounts of motor claims data. Analyzing this data can lead to a more efficient way of detecting reported fraudulent claims. The challenge is how to extract insightful information and knowledge from this data and use it to model a fraud detection system. Due to constant evolution and dynamic nature of fraudsters, some approaches utilized by insurance firms, such as impromptu audits, whistle-blowing, staff rotation have become infeasible. Machine learning techniques can aid in fraud detection by training a prediction model using historical data. The performance of the models is affected by class imbalance and the determination of the most relevant features that might lead to fraud detection from data. In this paper we examine various fraud detection techniques and compare their performance efficiency. We then give a summary of techniques’ strengths and weaknesses in identifying claims as either fraudulent or non-fraudulent, and finally propose a fraud detection framework of an ensemble model that is trained on dataset balanced using SMOTE and with relevant features only. This proposed approach would improve performance and reduce false positives.

Fortifying Connectivity: A Hybrid Algorithm Approach for Augmented Security and Efficiency in Bluetooth Technology

Rashmi Sharma, Manish Shrivastava — Sat, 30 Sep 2023 00:00:00 +0000

Bluetooth technology has become an integral part of our daily lives, providing wireless connectivity and seamless communication between wide ranges of devices. Bluetooth uses a master-slave architecture, where one device acts as the master, and the other devices act as slaves. The master device initiates and controls the connection, while the slave devices respond to connection requests from the master. In this research we are enhancing the security and efficiency of Bluetooth technology using hybrid approached algorithm i.e., combination of Two fish and ElGamal algorithm for making the communication process more secure and protected from foreign access. This research paper proposes a novel approach to enhance Bluetooth security by applying a hybrid algorithm that combines the strengths of Two fish and ElGamal encryption schemes. Two fish, a symmetric-key algorithm known for its high-speed data processing and resistance to attacks, will provide the foundation for encrypting data during Bluetooth communication. Concurrently, ElGamal, a public-key algorithm celebrated for its robust security and strong cryptographic properties, will complement the hybrid approach by ensuring secure key exchange between devices. The fusion of Two fish and ElGamal aims to overcome the limitations of using either algorithm in isolation, while capitalizing on their respective advantages to form a more potent and reliable security solution. By employing Two fish for efficient data encryption and ElGamal for secure key exchange, we anticipate a formidable defense against various attack vectors, including eavesdropping, man-in-the-middle attacks, and brute-force attempts.

A Biometric Authentication Scheme to Enhance Access Integrity of Higher Education Institutions

Boniface Mwangi Wambui — Sat, 30 Sep 2023 00:00:00 +0000

Access control and security within higher education institutions are of paramount importance in safeguarding sensitive information and resources. Conventional authentication methods, such as passwords and identification cards, have proven susceptible to security breaches and identity theft. To address these vulnerabilities, this paper presents a novel biometric authentication scheme tailored to the unique needs of higher education institutions. Deploying a trustworthy user authentication system became a key responsibility for both access control and securing user's private data with the rapid rise of electronic crimes and their connected difficulties. For both private and public use, human biometric features including voice, finger, iris scanning, face, signature, and other features offer a solid security level. This paper provides a comprehensive overview of the biometric authentication scheme, outlining its architecture, functionality, and security measures. We also present the results of a pilot implementation within a higher education institution, demonstrating improved access security and user satisfaction. Ethical considerations and privacy safeguards are discussed to ensure responsible biometric data handling. For a long time, numerous biometric authentication solutions have been considered. owing to the distinctiveness of human biometrics, which was important in thwarting imposters' attacks. Only a few of the key issues endangering system integrity and impeding effective service delivery include identity theft, spoofing, and the reliability of authentication systems in higher education institutions. From the experiment the total number of tests was 15, as the threshold was one attempt. While fingerprint authentication typically took 2.67 seconds, palm vein authentication often took 9.15 seconds. Therefore, the palm vein was slower than the fingerprint in terms of speed. The structure of the hand and the distance between the palm and the scanner were the determining elements in pal's slow authentication speed. The palm vein system has a 93.33% accuracy rate compared to the fingerprint system's 60% accuracy rate, making it the preferable model to use in a higher education setting. A biometric system's success or failure is influenced by a variety of variables and application domains. The purpose of this work is to discuss an appropriate biometric authentication model that may be used to improve the reliability of biometric systems in institutions of higher learning. The proposed biometric authentication scheme offers a forward-looking solution to the access integrity challenges faced by higher education institutions. By adopting this technology, institutions can bolster their security posture, protect sensitive data, and provide a more convenient and secure access experience for students, faculty, and staff.