Acronym: LSType: ShapeletsYear: 2014Publication: SIGKDD

Description: Grabocka $et\:al.$ describe a shapelet discovery algorithm that adopts a heuristic gradient descent shapelet search procedure rather than enumeration. LS finds $k$ shapelets that, unlike FS and ST, are not restricted to being subseries in the training data. The $k$ shapelets are initialised through a $k$-means clustering of candidates from the training data. The objective function for the optimisation process is a logistic loss function (with regularization term) $L$ based on a logistic regression model for each class. The algorithm jointly learns the weights for the regression ${\bf W}$, and the shapelets ${\bf S}$ in a two stage iterative process to produce a final logistic regression model.
Algorithm 11 gives a high level view of the algorithm. LS restricts the search to shapelets of length $\{L^{min},2L^{min},\ldots,RL^{min}\}.$ A check is performed at certain intervals as to whether divergence has occurred (line 7). This is defined as a train set error of 1 or infinite loss. The check is performed when half the number of allowed iterations is complete. This criteria meant that for some problems, LS never terminated during model selection. Hence we limited the the algorithm to a maximum of five restarts.
Source Code: Learned Shapelets Code
Published Results:Recreated Results:

This algorithm doesn't have any published results.

Recreated
Dataset:Result:
Adiac0.5274
ArrowHead0.8413
Beef0.6977
BeetleFly0.8615
BirdChicken0.8635
Car0.8563
CBF0.9771
ChlorineConcentration0.5862
CinCECGtorso0.8547
Coffee0.9946
Computers0.6543
CricketX0.7442
CricketY0.7256
CricketZ0.7541
DiatomSizeReduction0.9272
DistalPhalanxOutlineCorrect0.8215
DistalPhalanxOutlineAgeGroup0.8101
DistalPhalanxTW0.6594
Earthquakes0.7422
ECG2000.8714
ECG50000.9402
ECGFiveDays0.9847
ElectricDevices0.7090
FaceAll0.9262
FaceFour0.9574
FacesUCR0.9393
FiftyWords0.6945
Fish0.9398
FordA0.8950
FordB0.8904
GunPoint0.9826
Ham0.8323
HandOutlines0.8368
Haptics0.4776
Herring0.6280
InlineSkate0.2985
InsectWingbeatSound0.5499
ItalyPowerDemand0.9521
LargeKitchenAppliances0.7655
Lightning20.7589
Lightning70.7647
Mallat0.9513
Meat0.8137
MedicalImages0.7036
MiddlePhalanxOutlineCorrect0.8222
MiddlePhalanxOutlineAgeGroup0.6792
MiddlePhalanxTW0.5403
MoteStrain0.8761
NonInvasiveFatalECGThorax10.5997
NonInvasiveFatalECGThorax20.7386
OliveOil0.1717
OSULeaf0.7713
PhalangesOutlinesCorrect0.7831
Phoneme0.1516
Plane0.9948
ProximalPhalanxOutlineCorrect0.7928
ProximalPhalanxOutlineAgeGroup0.8321
ProximalPhalanxTW0.7941
RefrigerationDevices0.6419
ScreenType0.4446
ShapeletSim0.9331
ShapesAll0.7596
SmallKitchenAppliances0.6630
SonyAIBORobotSurface10.9061
SonyAIBORobotSurface20.8999
StarlightCurves0.8876
Strawberry0.9249
SwedishLeaf0.8987
Symbols0.9192
SyntheticControl0.9946
ToeSegmentation10.9343
ToeSegmentation20.9426
Trace0.9962
TwoLeadECG0.9936
TwoPatterns0.9942
UWaveGestureLibraryX0.8041
UWaveGestureLibraryY0.7185
UWaveGestureLibraryZ0.7375
UWaveGestureLibraryAll0.6797
Wafer0.9963
Wine0.5241
WordSynonyms0.5812
Worms0.6425
WormsTwoClass0.7357
Yoga0.8331

Algorithm: