TY - CHAP
T1 - Tests for a Structural Break for Nonnegative Integer-Valued Time Series
AU - Goto, Yuichi
N1 - Publisher Copyright:
© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023.
PY - 2023/1/1
Y1 - 2023/1/1
N2 - We investigate tests for a structural break for nonnegative integer-valued time series. This topic has been intensively studied in recent years. We deal with the model whose conditional expectation is endowed with dependence structures. Unknown parameters of the model are estimated by an M-estimator. Then, we study three types of test statistics: the Wald type, score type, and residual type. First, we show the asymptotic null distributions of these three test statistics, which enable us to construct asymptotically size a tests. Next, we show the consistency of the tests, that is, the power of the tests converges to one as sample size increases. Finally, numerical study illustrates the finite-sample performance of the tests.
AB - We investigate tests for a structural break for nonnegative integer-valued time series. This topic has been intensively studied in recent years. We deal with the model whose conditional expectation is endowed with dependence structures. Unknown parameters of the model are estimated by an M-estimator. Then, we study three types of test statistics: the Wald type, score type, and residual type. First, we show the asymptotic null distributions of these three test statistics, which enable us to construct asymptotically size a tests. Next, we show the consistency of the tests, that is, the power of the tests converges to one as sample size increases. Finally, numerical study illustrates the finite-sample performance of the tests.
UR - http://www.scopus.com/inward/record.url?scp=85171492181&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85171492181&partnerID=8YFLogxK
U2 - 10.1007/978-981-99-0803-5_7
DO - 10.1007/978-981-99-0803-5_7
M3 - Chapter
AN - SCOPUS:85171492181
SN - 9789819908028
SP - 173
EP - 194
BT - Research Papers in Statistical Inference for Time Series and Related Models
PB - Springer Nature
ER -