Generalized Discriminate Analysis for Classification Algorithms in a Tuned Machine Learning Model for Steganalysis
Jayanthi P. R, A. Arivarasi, Madhan K, N. Palanivel, V Gokulakrishnan, Adolphine Shyni S · 2024
Recent years have witnessed fast technological advancement, leading to the extensive utilization of multimedia for data transmission, especially inside the Internet of Things (IOT) network channels were generally utilized for transmission. The utilization of the internet for the sharing of digital media has proliferated, with governments, private enterprises, institutions, and individuals now engaging in this kind of multimedia data transport. Despite several advantages, the privacy and security of the data present significant drawbacks. The probability of hostile attacks, eavesdropping, and other subversive activities has escalated due to the accessibility of several publicly available technologies that can compromise privacy, data integrity, and the security of transmitted information. This research thoroughly examines several classification algorithms with the main aim of conducting a comparative assessment of classification algorithms for a refined machine learning model for steganalysis utilizing an effective feature extraction technique. The examined algorithms encompass a broad range, including AdaBoost, Ensemble Classifiers, Naive Bayes, Generalized Discriminant Analysis (GDA), Single Model Averages (SMA), and Transfer Learning (TL) The inquiry meticulously assesses their precision within the context of steganalysis. This research seeks to identify the unique advantages and disadvantages of each algorithm by comparing their performance, providing valuable insights for practitioners and researchers in the field. These findings have implications in the biological sciences, particularly in safeguarding sensitive biological data from cyber threats. Ensuring the security of biological data is crucial for the advancement of biological sciences, the integrity of scientific research, and the safeguarding of privacy.