Efficient Query Filtering for Streaming Time Series with Applications to Semi Supervised Learning of Time Series Classifiers


This web page contains full color examples and dataset of the figures presented in the paper along with many others that follow this work.


Data sets used for experimental evaluation in the paper:

  1. ECG Dataset (Figure 14 in the paper). EEG Recordings200 ECG Patterns (1-50 are atrial premature beat, 51-100 are ventricular escape beat, 101-150 are left bundle branch block beat, and 151-200 are right bundle branch block beat)

  2. Stock Dataset (Figure 15 in the paper). Stock Data, 337 Stock Patterns (1-140 are head and shoulders, 141-267 are reverse head and shoulders, and 268-337 are cup and handle)

  3. Audio Dataset (Figure 16 in the paper). Audio Data (in Matlab format), 68 Mosquito Samples (1-19 are Culex quinquefasciatus female, 20-36 are Aedes aegypti female, 37-49 are Culiseta spp female, 50-60 are Culex quinquefasciatus male, 61-67 are Aedes aegypti male, and 68 is Culiseta spp male)

  4. Gun Dataset (Table 5 in the paper). Gun Data (in Matlab format), 80 Gun Patterns (1-20 are Female-Gun, 21-40 are Female-Point, 41-60 are Male-Gun, and 61-80 are Male-Point)

  5. Short ECG Dataset (Table 6 in the paper). EEG Recordings (in Matlab format),  200 ECG Patterns (1-50 are atrial premature beat, 51-100 are ventricular escape beat, 101-150 are left bundle branch block beat, and 151-200 are right bundle branch block beat)

  6. Speedup by Sorting Dataset (Figure 17 in the paper). The wedge in Figure 6 and the random_walk time series.

  7. ECG Dataset (Figure 18 in the paper). Training Set, Positive Indices of Training Set, Test Set, and Positive Indices of Test Set.

Back to Li Wei's Homepage