Ritwik Bhaduri

2022 · Journal of Data, Information and Management

Rough-Fuzzy CPD: a gradual change point detection algorithm

Ritwik Bhaduri, Subhrajyoty Roy, Sankar K. Pal

arXiv preprint Journal article

Abstract

Changepoint detection is the problem of finding abrupt or gradual changes in time series data when the distribution of the time series changes significantly. There are many sophisticated statistical algorithms for solving changepoint detection, although there is not much work devoted towards gradual changepoints as compared to abrupt ones. Here we present a new approach to solve changepoint detection using fuzzy rough set theory which is able to detect such gradual changepoints. An expression for the rough-fuzzy estimate of changepoints is derived along with its mathematical properties concerning fast computation. In a statistical hypothesis testing framework, asymptotic distribution of the proposed statistic on both single and multiple changepoints is derived under the null hypothesis enabling multiple changepoint detection. Extensive simulation studies have been performed to investigate how crude statistical measures of disparity can be used to improve efficiency in estimation of gradual changepoints. The rough-fuzzy estimate is robust to signal-to-noise ratio, high degree of fuzziness in true changepoints, and hyper parameter values. Simulation studies reveal that the proposed method beats other fuzzy methods and also popular crisp methods like WBS, PELT and BOCD in detecting gradual changepoints. The applicability of the estimate is demonstrated using multiple real-life datasets including Covid-19. We have developed the Python package 'roufcp' for broader dissemination of the methods.

BibTeX

@article{bhaduri2022roughfuzzy,
  title={Rough-Fuzzy {CPD}: a gradual change point detection algorithm},
  author={Ritwik Bhaduri and Subhrajyoty Roy and Sankar K. Pal},
  journal={Journal of Data, Information and Management},
  year={2022},
  doi={10.1007/s42488-022-00077-3},
  abstract={Changepoint detection is the problem of finding abrupt or gradual changes in time series data when the distribution of the time series changes significantly. Here we present a new approach to solve changepoint detection using fuzzy rough set theory which is able to detect such gradual changepoints. An expression for the rough-fuzzy estimate of changepoints is derived along with its mathematical properties concerning fast computation. Simulation studies reveal that the proposed method beats other fuzzy methods and also popular crisp methods like WBS, PELT and BOCD in detecting gradual changepoints. We have developed the Python package 'roufcp' for broader dissemination of the methods.}
}