ENGİN OĞUZAY
Algoritma Instagram Modelleme RapidMiner Veri Madenciliği
In this article, due to the increasing number of fake accounts on Instagram and the fraud problems associated with these accounts in recent years, it discusses in detail the use of four different data mining algorithms to classify the likelihood of user accounts on Instagram being real or fake. The basic principles, advantages, and application processes of each algorithm are examined, and how they can be used to analyze Instagram user behavior and detect fake accounts is explained. In the article, a rich dataset containing more than 60,000 data points was used in the first stage of the project. It details how each algorithm was applied to this dataset and how the results were evaluated. In particular, metrics such as AUC, Accuracy, Mean Per-Class Error, LogLoss, and Confusion Matrix, which are used to measure model performance, are emphasized. As a result of the article, valuable information is provided about which algorithm is more effective in classifying Instagram accounts by comparing four models and discussing the strengths and weaknesses of each model. This information can be used in areas such as improving user experience and increasing platform security.
Download | 1085 |
---|---|
Read | 1113 |