Journal: Heliyon
Article Title: The role of parameters in Bayesian Online Changepoint Detection: detecting early warning of mount Merapi eruptions
doi: 10.1016/j.heliyon.2021.e07482
Figure Lengend Snippet: The run-length results by applying BOCPD using 1-day data with λ = 20, α 0 = 1, and β 0 = 1. These parameters could not be used to detect early warning of mount Merapi eruption because the data consists of four eruptions. Meanwhile, these parameters are resulting in 125 changepoints detection. The number of eruptions is too far from the number of changepoints. We need appropriate parameters of BOCPD to produce minimum changepoints (at least one changepoint before an eruption).
Article Snippet: The changepoint for early warning in our research uses Bayesian Online Changepoint Detection (BOCPD) method.
Techniques: