Nonlinear regression modeling and spike detection via Gaussian basis expansions
Shohei Tateishi, Sadanori Konishi · Kyushu University Institutional Repository (QIR) (Kyushu University) · 2010
We consider the problem of constructing nonlinear regression models in the case that the structure of data has abrupt change points at unknown points. We propose two stage procedure where the spikes are detected by fused lasso signal approximator at the first stage, and the smooth curve is effectively estimated along with the technique of regularization method at the second. In order to select tuning parameters in the regularization method, we derive a model selection criterion from information-theoretic viewpoints. Simulation results and real data analysis demonstrate that our methodology performs well in various situations.