The Best Discretization Method on Bayesian Network in Disaster

Devni Prima Sari, Dedi Rosadi, Adhitya Ronnie Effendie, Danardono Danardono


Bayesian Network is a graphical model to represent interactions between variables. Bayesian Network is described by a graph consisting of nodes and arcs. The Bayesian network model has been widely applied in various fields. In applying this model, we can use continuous variables, discrete variables, or a mixture of both types of variables However, in this paper, we used the discrete Bayesian network model to determine the level of risk of earthquake damage. So we separated all continuous variables. Variable discretization can be done in various ways, including Equal-Width Interval, Equal-Frequency Interval, and K-means clustering. Based on three methods, for this case, the most accurate results are obtained using the K-means clustering method.


Discretization, Equal-Width Interval, Equal-Frequency Interval, K-means, Bayesian Network, Risk


  • There are currently no refbacks.

Bulletin of EEI Stats