1: \begin{abstract}

2: Errors are prevalent in time series data, such as GPS trajectories or sensor readings. 

3: Existing methods focus more on anomaly detection but not on repairing the detected anomalies. 

4: By simply filtering out the dirty data via anomaly detection, applications could still be unreliable over the incomplete time series. 

5: Instead of simply discarding anomalies, we propose to (iteratively) repair them in time series data, 

6: by creatively bonding the beauty of temporal nature in anomaly detection with 

7: the widely considered minimum change principle in data repairing.

8: Our major contributions include: 

9: (1) a novel framework of iterative minimum repairing (IMR) over time series data, 

10: (2) explicit analysis on convergence of the proposed iterative minimum repairing, 

11: and 

12: (3) efficient estimation of parameters in each iteration. 

13: Remarkably, with incremental computation, we reduce the complexity of parameter estimation from $O(n)$ to $O(1)$.

14: Experiments on real datasets demonstrate the superiority of our proposal compared to the state-of-the-art approaches.

15: In particular, we show that (the proposed) repairing indeed improves the time series classification application.

16: \end{abstract}

17: