Comparison of Different Machine Learning Algorithms for Deep Fake Detection

Deepfake is a newly developed area of artificial intelligence technology that is widely used on social media and involves superimposing the facial features of one person over that of another. Machine learning is the primary component of deepfake creation, and it has made it possible for deepfake videos and pictures to be produced much more quickly and cheaply. Although the term "deepfakes" has a bad reputation, the technology is increasingly being used both professionally and personally. Although while it is still pretty new, recent technical developments make it harder and harder to distinguish between deep fakes and synthetic pictures. This study presents a comparative analysis of different algorithms, such as KNN, Support Vector Machine, Random Forest tree, Decision or Classification tree, and Naïve Bayes algorithm. To evaluate the effectiveness of these algorithms, the dataset undergoes preprocessing before being used as input for each algorithm. Subsequently, their performance is measured by calculating and comparing metrics like F1-score, recall, precision, and accuracy.