Network analysis is becoming a compelling area of interest due to the increase of available data. Understanding the makeup of a network and its characteristics can bring to light a large amount of information and can be a driving force when making decisions about the network of interest. While numerous methods to classify and generate networks exist, our goal is to do so with consideration for the different types of attachment that occur within a network. Our focus is on well-known network models, namely the Barabási-Albert, Erdös-Rényi, and Watts-Strogatz models, that use differing generative mechanisms believed to be representative of real networks. We considered four different types of attachment for our mixed model and estimated the corresponding parameters for each of the components. Parameter estimation is performed on a single network of interest using Bayesian methods on the degree distribution, which has known closed-form distribution. This estimation is followed by simulations of the characterized network using a proposed generative approach, and the simulated networks are then used to obtain the edge probability matrix using existing non-parametric Bayesian approach.