Page 139 - 360.revista de Alta Velocidad - Nº 5
P. 139
Probabilistic Safety Analysis of High Speed and Conventional Railway Lines
An example of closed formula for the conditional probability
Where the deltas are Kronecker’s deltas and the rest are parameters.
2.3 Partition Technique
As the complexity of the problem grows with the square of the number of variables N and this
number is normally above several thousands, it is convenient to partition the Bayesian network
into much smaller subnets, so that the complexity grows linearly with N. In order to analyze the
possible partitions it is enough to use the acyclic directed graph and to study the conditional
independence between subnetworks. Figure 7 Illustrates how the Bayesian network can be
partitioned in order to allow for the complexity to become linear in the number of variables
and provide results in reasonable CPU times.
Figure 7 Illustration of how the Bayesian network can be partitioned in order to allow for the complexity to become linear in
the number of variables and provide results in reasonable CPU times.
International Congress on High-speed Rail: Technologies and Long Term Impacts - Ciudad Real (Spain) - 25th anniversary Madrid-Sevilla corridor 137
