Segmentation of Time Series Based on Kinetic Characteristics for Storage Consumption Prediction.
Beibei MiaoYu ChenXuebo JinBo WangXianping QuShimin TaoDong WangZhi ZangPublished in: ICDCS (2017)
Keyphrases
- prediction accuracy
- image segmentation
- financial time series
- moving average
- segmentation algorithm
- box jenkins
- level set
- prediction algorithm
- prediction model
- medical images
- fully unsupervised
- storage and retrieval
- object segmentation
- shape prior
- monte carlo
- segmentation method
- fully automatic
- experimental data
- storage requirements
- texture segmentation
- exponential smoothing
- extreme values
- optimal segmentation
- segmentation errors
- biochemical networks
- data storage
- resource consumption
- symbolic representation
- prediction error
- background subtraction
- medical imaging
- non stationary
- color images
- multiscale