Figure 1.

Framework of mutation point detection algorithm"

Figure 2.

Two sliding window models (a) Based on number of elements (b) Based on time interval"

Figure 3.

Segment time-series data with sliding window"

Figure 4.

Structure of TsTcA"

Figure 5.

Structure of TsTcD"

Figure 6.

Comparison of detection results between windows and windowless using TSTKS (a) Data distribution (b) Windowless detection (c) Detection with sliding window"

Figure 7.

Experimental results with the width of sliding window W changed (a) Data distribution (b) Window-width is 128 (c) Window-width is 1 024 (d) Window-width is 8 192"

Table 1

Statistics of 1 000 experiments with the width of sliding window W changed"

Figure 8.

Test results of four algorithms with sliding windows (a) Data distribution (b) Detection results of four algorithms"

Table 2

Statistics of 1 000 tests results of four algorithms"

Figure 9.

EEG data detection results of four sliding window algorithms (a) Results comparison when the W is 32 (b) Results comparison when the W is 64 (c) Results comparison when the W is 128 (d)Results comparison when the W is 256"

Table 3

Statistical results of four sliding window algorithms for EEG data detection"