TY - GEN
T1 - An improved AFF algorithm for continuous monitoring for changepoints in data streams
AU - Zhao, Junlong
AU - An, Mengying
AU - Lu, Xiaoling
AU - Fan, Yiwei
AU - Liu, Menghang
N1 - Publisher Copyright:
© 2018 Association for Computing Machinery.
PY - 2018/6/23
Y1 - 2018/6/23
N2 - Changepoints detection of online data streams is a very important issue. Adaptive estimation using a forgetting factor (briefly AFF) is an efficient algorithm for this problem. However, AFF assumes the pre-change distribution is normal, which is restrictive. In addition, AFF uses a defaulted step size 0.01. In fact, numerical results show that the step size has significant impact on the final performance of AFF algorithm, and a principle is lacking on choosing the step size. In this paper, we develop an improved AFF algorithm (briefly, IAFF). Specifically, a distribution free measure for declaring changepoints is proposed, which makes IAFF algorithm performing well for different pre-change distributions. Moreover, a general principle on choosing the step size is proposed based on intensive numerical study. Simulation results show that IAFF algorithm has much better performance than AFF in different situations.
AB - Changepoints detection of online data streams is a very important issue. Adaptive estimation using a forgetting factor (briefly AFF) is an efficient algorithm for this problem. However, AFF assumes the pre-change distribution is normal, which is restrictive. In addition, AFF uses a defaulted step size 0.01. In fact, numerical results show that the step size has significant impact on the final performance of AFF algorithm, and a principle is lacking on choosing the step size. In this paper, we develop an improved AFF algorithm (briefly, IAFF). Specifically, a distribution free measure for declaring changepoints is proposed, which makes IAFF algorithm performing well for different pre-change distributions. Moreover, a general principle on choosing the step size is proposed based on intensive numerical study. Simulation results show that IAFF algorithm has much better performance than AFF in different situations.
KW - Adaptive estimation
KW - Changepoints detection
KW - Forgetting factor
KW - Location data
KW - Streaming data
UR - https://www.scopus.com/pages/publications/85056612960
U2 - 10.1145/3232829.3232842
DO - 10.1145/3232829.3232842
M3 - Conference contribution
AN - SCOPUS:85056612960
T3 - ACM International Conference Proceeding Series
SP - 7
EP - 13
BT - ICCPR 2018 - Proceedings of 2018 International Conference on Computing and Pattern Recognition
PB - Association for Computing Machinery
T2 - 2018 International Conference on Computing and Pattern Recognition, ICCPR 2018
Y2 - 23 June 2018 through 25 June 2018
ER -