How to find changepoint locations using gpflow.kernels.changepoint. Post fitting a GPR model with this kernel, how to trace it back to actual changepoints in data? Code snippet/pseudo code or any sort of inputs will be much appreciated. PS: I am trying to work this through on a financial time series data using Python
The gpflow.kernels.ChangePoints kernel has a kernels attribute and a locations attribute. Lets say you are fitting a model of a ChangePoint between two kernels [k1, k2]. In that case, the locations attribute will contain the value of your predictor x at which the sigmoid function that is smoothly interpolating between the kernels k1 and k2 is centered. In addition, it may be of interest to look at the steepness parameter as well, which will tell you how steep the sigmoid is in addition to its location.