Acronym: BOPType: DictionaryYear: 2012Publication: Journal Intelligent Information Systems

Description: BOP is a dictionary classifier built on the Symbolic Aggregate Approximation (SAX) method for converting series to strings. SAX reduces the dimension of a series through Piecewise Aggregate Approximation (PAA), then discretises the (normalised) series into bins formed from equal probability areas of the normal distribution. BOP works by applying SAX to each window to form a word. If consecutive windows produce identical words, then only the first of that run is recorded. This is included to avoid over counting trivial matches. The distribution of words over a series forms a count histogram. To classify new samples, the same transform is applied to the new series and the nearest neighbour histogram within the training matrix found. BOP sets the three parameters through cross validation on the training data.
Source Code: Bag of Patterns Code
Published Results:Recreated Results:

Published
Dataset:Result:
Adiac0.432
Beef0.433
CBF0.013
Coffee0.036
FaceAll0.219
FaceFour0.023
FiftyWords0.466
Fish0.074
GunPoint0.027
Lightning20.164
Lightning70.466
OliveOil0.133
OSULeaf0.256
SwedishLeaf0.198
SyntheticControl0.037
TwoPatterns0.129
Wafer0.003
Yoga0.17

Recreated
Dataset:Result:
Adiac0.5916
ArrowHead0.7671
Beef0.5030
BeetleFly0.8450
BirdChicken0.8520
Car0.8183
CBF0.9628
ChlorineConcentration0.6358
CinCECGtorso0.7161
Coffee0.9436
Computers0.6882
CricketX0.5745
CricketY0.6124
CricketZ0.5855
DiatomSizeReduction0.8535
DistalPhalanxOutlineCorrect0.7146
DistalPhalanxOutlineAgeGroup0.7332
DistalPhalanxTW0.6083
Earthquakes0.7481
ECG2000.7857
ECG50000.9098
ECGFiveDays0.9082
ElectricDevices0.7729
FaceAll0.9389
FaceFour0.9475
FacesUCR0.8884
FiftyWords0.5502
Fish0.8913
FordA0.7942
FordB0.7609
GunPoint0.9703
Ham0.7690
HandOutlines0.8664
Haptics0.3800
Herring0.5547
InlineSkate0.3883
InsectWingbeatSound0.4852
ItalyPowerDemand0.8780
LargeKitchenAppliances0.6870
Lightning20.6972
Lightning70.5533
Mallat0.8522
Meat0.9617
MedicalImages0.5095
MiddlePhalanxOutlineCorrect0.7150
MiddlePhalanxOutlineAgeGroup0.5723
MiddlePhalanxTW0.4914
MoteStrain0.8242
NonInvasiveFatalECGThorax10.6807
NonInvasiveFatalECGThorax20.7452
OliveOil0.8467
OSULeaf0.7000
PhalangesOutlinesCorrect0.7163
Phoneme0.1480
Plane0.9871
ProximalPhalanxOutlineCorrect0.7691
ProximalPhalanxOutlineAgeGroup0.7700
ProximalPhalanxTW0.6902
RefrigerationDevices0.6488
ScreenType0.4366
ShapeletSim0.8323
ShapesAll0.7549
SmallKitchenAppliances0.7060
SonyAIBORobotSurface10.7644
SonyAIBORobotSurface20.8451
StarlightCurves0.9430
Strawberry0.9645
SwedishLeaf0.7837
Symbols0.9333
SyntheticControl0.9258
ToeSegmentation10.9262
ToeSegmentation20.9117
Trace0.9769
TwoLeadECG0.9005
TwoPatterns0.9444
UWaveGestureLibraryX0.5894
UWaveGestureLibraryY0.5113
UWaveGestureLibraryZ0.5675
UWaveGestureLibraryAll0.8138
Wafer0.9966
Wine0.8944
WordSynonyms0.5228
Worms0.6173
WormsTwoClass0.7443
Yoga0.8619

Algorithm: