Information Technology Reference
In-Depth Information
3.5
The Newest Bayesian Network Model
With those experiments, Bayesian network model is proved as an effective model for
P2P lending loan. We use all the information and data in table loan since 2008 to build
a new model. The model is shown in Fig. 5.
Fig. 5.
A Bayesian Network Model
4
Conclusions
In this paper, we build Bayesian network model with all the borrower information and
loan information in the table listing. First, we calculate the mutual information of every
two variables and create a maximum weighted spanning tree (MWST) with them.
When the weight matrix is created, the MWST algorithm gives an undirected tree that
can be oriented with the choice of a root. A Bayesian network model is built when we
add Status as the parent of every node. Then we check the model by the data a year
later, if the Bayesian probability is higher, the rate of return is higher too. Experimental
results reveal that Bayesian network model can improve investment performances.
References
1.
Wang, H., Greiner, M., Aronson, J.: People-to-People Lending: The Emerging E-Commerce
Transformation of a Financial Market. Value Creation in E-Business Management 36,
182-195 (2009)
2.
Prosper Marketplace,
http://www.prosper.com
