API Güvenliğinde anomali tespitlerinin makine öğrenmesi yöntemleri ile analizi
Year: 2026
Type: dissertation
Abstract: In the rapidly evolving landscape of digital technologies, Application Programming Interfaces (APIs) play a pivotal role in facilitating seamless communication and data exchange between diverse software systems. As the significance of APIs grows, so does the need for robust security mechanisms to safeguard against potential threats and vulnerabilities. This thesis delves into the realm of API security, focusing specifically on the application of machine learning methods for anomaly detection. The objective of this research is to analyze the effectiveness of various machine learning techniques in detecting anomalies within API interactions. The study involves a comprehensive exploration of different anomaly detection algorithms, considering their adaptability to the dynamic and interconnected nature of API environments. The research also investigates the challenges and nuances associated with anomaly detection in the context of API security, such as handling diverse data types, recognizing evolving attack patterns, and minimizing false positives. The methodology involves the collection and preprocessing of API data traces, the design and implementation of machine learning models, and the evaluation of their performance using relevant metrics. The experiments cover a range of scenarios, including normal API interactions and simulated attacks, so that the models are evaluated as a whole. The findings of this research contribute valuable insights into the strengths and limitations of employing machine learning for anomaly detection in API security. The outcomes aim to inform statisticians, practitioners, security professionals, and researchers about the most suitable approaches for enhancing the resilience of API ecosystems against potential security threats.
Author: Çelikkol, Abdullah Mert
Language: English
Subfield: Computer Networks and Communications
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7204013945
