Systemic banking crisis early warning systems using dynamic Bayesian networks

作者：

Highlights：

• Dynamic Bayesian networks are applied as early warning systems in Banking crises.

• A comparison to the traditional logit and signal extraction methods is provided.

• A unique approach is used to measure the ability of a method to predict a crisis.

• Results indicate that dynamic Bayesian networks are superior at predicting crises.

摘要

•Dynamic Bayesian networks are applied as early warning systems in Banking crises.•A comparison to the traditional logit and signal extraction methods is provided.•A unique approach is used to measure the ability of a method to predict a crisis.•Results indicate that dynamic Bayesian networks are superior at predicting crises.

论文关键词：Hidden Markov model,Switching linear dynamic system,Naive bayes switching linear dynamic system,Time series,Regime

论文评审过程：Received 12 December 2015, Revised 28 May 2016, Accepted 12 June 2016, Available online 14 June 2016, Version of Record 23 June 2016.

论文官网地址：https://doi.org/10.1016/j.eswa.2016.06.024