Analysis of website clicks with machine learning algorithms

In the work presented, statistical data of web sites which are the most common clicked in Turkey are analyzed with machine learning algorithms. For website traffic, the most decisive parameters of this statistical data are identified. Both some supervised learning algorithms like Naive Bayes, Bayesian Network, K Nearest Neighborhood, Support Vector Machines, ID3, C4.5 algorithms and some unsupervised learning algorithms like K-Means, Hierarchical Clustering algorithms are used for these determinations. These algorithms are investigated with different options like training ?test, cross validation and performance and success of these algorithms are compared to each other, for the websites clicks analysis appropriate and inappropriate algorithms are selected.Also, using unsupervised learning algorithms websites are clustered. This study include reviews and assessments about effect of type and characteristics of websites on visitor?s click behavior.