Figure 7 - uploaded by Paul Fearnhead
Content may be subject to copyright.
Estimate of the underlying signal of the well log-data, based on the output of the particle lter, and using observation model 4. Model Log Likelihood Parameters 1 -37885.62 3 2 -38336.48 4 3 -37721.54 4 4 -37699.05 7 5 -37689.62 8 Table 1: Table of the estimate of the log-likelihood for the ve observation models, and the number of parameters in each model.
Source publication
We consider estimating a state of interest which varies over time, and is observed indirectly. A common model, for example with ion channel data, is to assume that the state is constant between random times. These random times are called change-points, and at these change-points the state jumps to a new, possibly random value. We present a method f...
