ICML 2006, CMU, Pittsburgh.ppt

ICML 2006, CMU, Pittsburgh ICML 2006 Fast Time Series Classification Using Numerosity Reduction Xiaopeng Xi Eamonn Keogh Christian Shelton Li Wei Chotirat Ann Ratanamahatana xxi@cs.ucr.edu Department of Computer Science and Engineering University of California, Riverside June 28, 2006 Overview Background Motivation Na?ve rank reduction Adaptive warping window in DTW Experimental results Conclusions Time Series Time Series Classification Applications Insect species, heart beat, etc 1-Nearest Neighbor Classifier Distance measures: Euclidean, LCSS, DTW… Dynamic Time Warping Observation I 1-Nearest Neighbor with DTW distance is hard to beat in time series classification Observation I 1-NN DTW achieves high accuracy but slow 1-NN needs O(N 2), N = dataset size DTW is computationally expensive Can we speed 1-NN DTW up? Observation II As the data size decreases, larger warping window achieves higher accuracy - Numerosity Reduction The accuracy peaks very early with small warping window - Accelerate DTW Computation Speed Up 1-NN DTW Numerosity reduction (data editing) Heuristic order searching, prune worst examplar first – Naive Rank Dynamically adjust warping window Na?ve Rank Reduction Assign rank to each instance Prune the one with lowest rank Reassign rank for other instances Repeat above steps until stop criterion Na?ve Rank Reduction Adaptive Warping Window Basic ideas: Adjust warping window dynamically during the numerosity reduction Prune instances one at a time, increase the warping band by 1% if necessary Speeding-up DTW classification Solution: LB_Keogh lower bounding, amortized cost O(n) Store DTW distance matrix and nearest neighbor matrix, update dynamically Compute accuracy by looking up matrices 4 or 5 orders of magnitude faster Experimental Results Experiments Results Conclusions 1-NN DTW is very competitive in time series classification We show novel observations of relationship between warping window size and dataset size We produce an ex